You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.

command.h 7.9 kB

[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
adreno image shader + fp16 + fp16a (#1714) * wip * wip * fix * image and imageview can not be destroyed until command execution ends * fast copy path for tightly packed data * wip * texture load works * 1d 3d image * record clone image, multiple commands share one image reference * upload download image * layer forward accept vkimagemat * vkimagemat graph works * staging vkimagemat for passing dynamic parameters, macro for fp32+image shader, padding image shader * vkimagemat elemsize * convolution test pass * conv1x1s1 image shader * fast staging image allocator from host memory, pooling image shader * convolutiondepthwise image shader * innerproduct image shader * packing image shader * crop deconvolution image shader * resolve spirv binding types * image fp16 and fp16a, cast image shader * eltwise image shader * wip * absval image shader * deconvolutiondepthwise image shader * concat image shader, squeezenet works * noop split image shader * uniform precision hint * layer support_image_storage * wip * vulkan device utility operator * command is storage and packing option aware * fallback to cpu on image allocation failed, mobilenetssd works * flatten image shader, enable more test * ci test * check imgfp32 imgfp16 imgfp16a features * fix ci test * fix ci test * upgrade swiftshader * wip * opt aggressive * imgfp16p * opt none * convolution winograd image shader * fix flush range, fast copy path for continous buffer * minor fix * fix innerproduct * wip ... * wip * cast fix * packing test * wip * image fp16p is fp16p * wip * silence * more line info * code clean * softmax image shader
6 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
[WIP] vulkan compute (#618) * vulkan infrastructure * vkallocator and vkmat * layer interface for vulkan compute * wip... * default vulkan device, command wrapper, upload model weight in load_model to simplify layer interface * simplify command api, vkmat holds staging buffer, relu works * initialize specialization constant, simplify command dispatch, fix staging buffer copy with different shape, convolution works * init extension functions * dynamic local size and group count * group count=1 is invalid * regard device max workgroup size limit * fix relu oooops * decouple command record and staging allocation * create result blob * add pooling shader * buffer is faster than image :) * fix pooling shader * add innerproduct shader * readonly writeonly decoration * simplify buffer creation * decouple command and layer, VK_KHR_descriptor_update_template extension makes descriptor binding update easy :D * fix vulkan building issues in visual studio (#1) * fix building issues on visual studio * ignore benchmark * cancel changes * ... ... * decouple paramdict and vulkandevice * fix staging buffer destroy in model loading * remove vkdev member in option * add padding shader * simplify vulkan layer creation, simplify convolution and pooling shader for no padding, less debug output * add convolutiondepthwise and softmax shader * specialization float type, add leakyrelu * add dropout shader * add batchnorm shader * split vulkan forward * add scale shader * push constant type can be int or float * set_optimal_local_size_xyz * add eltwise shader * concat vulkan forward * fix convolution without bias * add dummy shader for concat and split, more fix ... * optional VK_KHR_descriptor_update_template and VK_KHR_push_descriptor * check VK_KHR_push_descriptor for vkCmdPushDescriptorSetWithTemplateKHR * binaryop and unaryop shader * hide raw command buffer * simple vkbenchncnn benchmark * create device with transfer queue * rename command to vkcompute, add vktransfer and layer upload_model interface * external VkMat, copy and map wrt buffer offset * command copy respect offset and size * decouple weight upload and load, simplify upload weight api, use one big staging buffer for uploading weights * fix build on android * binding count can not vary :( * barrier check state, fix sub-op destruction * declare local_size_xyz constant, fix crash on radv * fix local_size_xyz, second try * more barrier and state fix * fix softmax * reconstruct buffer memory allocator, reuse blob buffer, less verbose output * find unified memory type index * weight staging buffer allocator and weight buffer allocator, respect descriptor buffer offset alignment * use VK_KHR_descriptor_update_template for faster descriptor update if available, multithread pipeline creation * find more useful vulkan extensions and enable them * fix msvc build * respect VK_KHR_dedicated_allocation for weight buffer allocation * fix android build * fix bias name conflicts with metal * decouple pipeline and layer, building shader sources into shader module, dedicated create_pipeline api, simplify pipeline recording * drop dummy shader, inplace softmax, multiple shader module works * fix unique queue family index error * flatten support vulkan * mnasnet run * find shader module by name, each entry point per shader module, fix attribute/id conflict on moltenvk * some minor changes * add some high level api * use dedicated transfer queue to upload weight model * prefer mappable buffer on unified memory * global pooling and convolution fc, reuse staging buffer * implement ring-buffer style blob allocator, add VkBufferMemory capacity * use blob allocator for workspace blob, it works fine :) * vulkan option off * Update layer.cpp * fix build with vulkan off * less verbose output, fix crash on vulkan_compute off * merge benchncnn tool * allocator clear api, use new weight buffer allocator per net * add default locked allocator * mapped mat ptr api, persistent mapped memory works generally :) * travis ci linux vulkan * travis ci vulkan wip ... * more gpu wip ... * more gpu wip ... * wip... * wip... * wip... ... * wip... ios vulkan build... * find glslangValidator on ios build * use dynamic moltenvk library * travis ci wip ... * ios simulator does not support metal at all * fix cpu only extractor * optimize workgroup size, first try * optimize workgroup size, second try * conv1x1s1d1 vec4 * revert build system * fix ncnn2mem build * fix ncnn2mem build
7 years ago
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216
  1. // Tencent is pleased to support the open source community by making ncnn available.
  2. //
  3. // Copyright (C) 2020 THL A29 Limited, a Tencent company. All rights reserved.
  4. //
  5. // Licensed under the BSD 3-Clause License (the "License"); you may not use this file except
  6. // in compliance with the License. You may obtain a copy of the License at
  7. //
  8. // https://opensource.org/licenses/BSD-3-Clause
  9. //
  10. // Unless required by applicable law or agreed to in writing, software distributed
  11. // under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
  12. // CONDITIONS OF ANY KIND, either express or implied. See the License for the
  13. // specific language governing permissions and limitations under the License.
  14. #ifndef NCNN_COMMAND_H
  15. #define NCNN_COMMAND_H
  16. #include "platform.h"
  17. #if NCNN_VULKAN
  18. #include <vulkan/vulkan.h>
  19. #include "mat.h"
  20. namespace ncnn {
  21. class Pipeline;
  22. class VkCompute
  23. {
  24. public:
  25. VkCompute(const VulkanDevice* vkdev);
  26. virtual ~VkCompute();
  27. public:
  28. void record_upload(const Mat& src, VkMat& dst, const Option& opt);
  29. void record_upload(const Mat& src, VkImageMat& dst, const Option& opt);
  30. void record_download(const VkMat& src, Mat& dst, const Option& opt);
  31. void record_download(const VkImageMat& src, Mat& dst, const Option& opt);
  32. void record_buffer_to_image(const VkMat& src, VkImageMat& dst, const Option& opt);
  33. void record_image_to_buffer(const VkImageMat& src, VkMat& dst, const Option& opt);
  34. void record_clone(const Mat& src, VkMat& dst, const Option& opt);
  35. void record_clone(const Mat& src, VkImageMat& dst, const Option& opt);
  36. void record_clone(const VkMat& src, Mat& dst, const Option& opt);
  37. void record_clone(const VkImageMat& src, Mat& dst, const Option& opt);
  38. void record_clone(const VkMat& src, VkMat& dst, const Option& opt);
  39. void record_clone(const VkImageMat& src, VkImageMat& dst, const Option& opt);
  40. void record_clone(const VkMat& src, VkImageMat& dst, const Option& opt);
  41. void record_clone(const VkImageMat& src, VkMat& dst, const Option& opt);
  42. void record_pipeline(const Pipeline* pipeline, const std::vector<VkMat>& bindings, const std::vector<vk_constant_type>& constants, const VkMat& dispatcher);
  43. void record_pipeline(const Pipeline* pipeline, const std::vector<VkImageMat>& bindings, const std::vector<vk_constant_type>& constants, const VkImageMat& dispatcher);
  44. void record_pipeline(const Pipeline* pipeline, const std::vector<VkMat>& buffer_bindings, const std::vector<VkImageMat>& image_bindings, const std::vector<vk_constant_type>& constants, const VkMat& dispatcher);
  45. void record_pipeline(const Pipeline* pipeline, const std::vector<VkMat>& buffer_bindings, const std::vector<VkImageMat>& image_bindings, const std::vector<vk_constant_type>& constants, const VkImageMat& dispatcher);
  46. void record_pipeline(const Pipeline* pipeline, const std::vector<VkMat>& buffer_bindings, const std::vector<VkImageMat>& image_bindings, const std::vector<vk_constant_type>& constants, const Mat& dispatcher);
  47. #if NCNN_BENCHMARK
  48. void record_write_timestamp(uint32_t query);
  49. #endif // NCNN_BENCHMARK
  50. #if __ANDROID_API__ >= 26
  51. void record_import_android_hardware_buffer(const ImportAndroidHardwareBufferPipeline* pipeline, const VkImageMat& src, const VkMat& dst);
  52. void record_import_android_hardware_buffer(const ImportAndroidHardwareBufferPipeline* pipeline, const VkImageMat& src, const VkImageMat& dst);
  53. #endif // __ANDROID_API__ >= 26
  54. int submit_and_wait();
  55. int reset();
  56. #if NCNN_BENCHMARK
  57. int create_query_pool(uint32_t query_count);
  58. int get_query_pool_results(uint32_t first_query, uint32_t query_count, std::vector<uint64_t>& results);
  59. #endif // NCNN_BENCHMARK
  60. protected:
  61. int init();
  62. int begin_command_buffer();
  63. int end_command_buffer();
  64. protected:
  65. const VulkanDevice* vkdev;
  66. VkCommandPool compute_command_pool;
  67. VkCommandBuffer compute_command_buffer;
  68. VkFence compute_command_fence;
  69. std::vector<VkMat> upload_staging_buffers;
  70. std::vector<VkMat> download_post_buffers;
  71. std::vector<Mat> download_post_mats_fp16;
  72. std::vector<Mat> download_post_mats;
  73. std::vector<VkImageMemory*> image_blocks_to_destroy;
  74. // the good-old path for device without VK_KHR_push_descriptor
  75. std::vector<VkDescriptorPool> descriptor_pools;
  76. std::vector<VkDescriptorSet> descriptorsets;
  77. struct record
  78. {
  79. enum
  80. {
  81. TYPE_copy_buffer,
  82. TYPE_copy_image,
  83. TYPE_copy_buffer_to_image,
  84. TYPE_copy_image_to_buffer,
  85. TYPE_bind_pipeline,
  86. TYPE_bind_descriptorsets,
  87. TYPE_push_constants,
  88. TYPE_dispatch,
  89. TYPE_memory_barrers,
  90. TYPE_buffer_barrers,
  91. TYPE_image_barrers,
  92. #if NCNN_BENCHMARK
  93. TYPE_write_timestamp,
  94. #endif // NCNN_BENCHMARK
  95. TYPE_post_download,
  96. TYPE_post_cast_float16_to_float32,
  97. };
  98. int type;
  99. VkCommandBuffer command_buffer;
  100. union
  101. {
  102. struct { VkBuffer src; VkBuffer dst; uint32_t region_count; const VkBufferCopy* regions; } copy_buffer;
  103. struct { VkImage src; VkImageLayout src_layout; VkImage dst; VkImageLayout dst_layout; uint32_t region_count; const VkImageCopy* regions; } copy_image;
  104. struct { VkBuffer src; VkImage dst; VkImageLayout layout; uint32_t region_count; const VkBufferImageCopy* regions; } copy_buffer_to_image;
  105. struct { VkImage src; VkImageLayout layout; VkBuffer dst; uint32_t region_count; const VkBufferImageCopy* regions; } copy_image_to_buffer;
  106. struct { VkPipelineBindPoint bind_point; VkPipeline pipeline; } bind_pipeline;
  107. struct { VkPipelineBindPoint bind_point; VkPipelineLayout pipeline_layout; uint32_t descriptorset_count; uint32_t descriptorset_offset; } bind_descriptorsets;
  108. struct { VkPipelineLayout pipeline_layout; VkShaderStageFlags stage_flags; uint32_t size; const void* values; } push_constants;
  109. struct { uint32_t group_count_x; uint32_t group_count_y; uint32_t group_count_z; } dispatch;
  110. struct { VkPipelineStageFlags src_stage; VkPipelineStageFlags dst_stage; uint32_t barrier_count; const VkMemoryBarrier* barriers; } memory_barrers;
  111. struct { VkPipelineStageFlags src_stage; VkPipelineStageFlags dst_stage; uint32_t barrier_count; const VkBufferMemoryBarrier* barriers; } buffer_barrers;
  112. struct { VkPipelineStageFlags src_stage; VkPipelineStageFlags dst_stage; uint32_t barrier_count; const VkImageMemoryBarrier* barriers; } image_barrers;
  113. #if NCNN_BENCHMARK
  114. struct { uint32_t query; } write_timestamp;
  115. #endif // NCNN_BENCHMARK
  116. struct { uint32_t download_post_buffer_mat_offset; uint32_t download_post_mat_fp16_offset; } post_download;
  117. struct { uint32_t download_post_mat_fp16_offset; uint32_t download_post_mat_offset; } post_cast_float16_to_float32;
  118. };
  119. };
  120. std::vector<record> delayed_records;
  121. #if NCNN_BENCHMARK
  122. uint32_t query_count;
  123. VkQueryPool query_pool;
  124. #endif // NCNN_BENCHMARK
  125. };
  126. class VkTransfer
  127. {
  128. public:
  129. VkTransfer(const VulkanDevice* vkdev);
  130. ~VkTransfer();
  131. public:
  132. void record_upload(const Mat& src, VkMat& dst, const Option& opt);
  133. void record_upload(const Mat& src, VkImageMat& dst, const Option& opt);
  134. int submit_and_wait();
  135. protected:
  136. int init();
  137. int begin_command_buffer();
  138. int end_command_buffer();
  139. protected:
  140. const VulkanDevice* vkdev;
  141. VkCommandPool compute_command_pool;
  142. VkCommandPool transfer_command_pool;
  143. VkCommandBuffer upload_command_buffer;
  144. VkCommandBuffer compute_command_buffer;
  145. VkSemaphore upload_compute_semaphore;
  146. VkFence upload_command_fence;
  147. VkFence compute_command_fence;
  148. std::vector<VkMat> upload_staging_buffers;
  149. };
  150. } // namespace ncnn
  151. #endif // NCNN_VULKAN
  152. #endif // NCNN_COMMAND_H