dora-rs
/
dora
Not watched
1
0
Fork 0
Code
Releases 124 Wiki evaluate Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain HPC
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.
2751 Commits
57 Branches
16 MB
23287 Download
 
 
 
 
 
 
Tree: b0e6a160fe
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'b0e6a160fe'
${ noResults }
dora/examples/cuda-benchmark
haixuanTao b0e6a160fe Small renaming of buffer to ipc buffer to make ipc buffer more explicit 1 year ago
..
README.md Small renaming of buffer to ipc buffer to make ipc buffer more explicit 1 year ago
cpu_bench.yml Small renaming of buffer to ipc buffer to make ipc buffer more explicit 1 year ago
cuda_bench.yml Small renaming of buffer to ipc buffer to make ipc buffer more explicit 1 year ago
cuda_receiver.py Small renaming of buffer to ipc buffer to make ipc buffer more explicit 1 year ago
cuda_sender.py Small renaming of buffer to ipc buffer to make ipc buffer more explicit 1 year ago
helper.py Small renaming of buffer to ipc buffer to make ipc buffer more explicit 1 year ago

README.md

Example of latency using CUDA Zero Copy instead of regular Shared Memory over CPU

Install

pip install dora-rs-cli # if not already present

# Install pyarrow with gpu support
conda install pyarrow "arrow-cpp-proc=*=cuda" -c conda-forge

# Install numba for translation from arrow to torch
pip install numba

# Install torch if it's not already present
pip install torch

Run

dora run cpu_bench.yml

dora run cuda_bench.yml

cat benchmark_data.csv
Delete task

Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.

cancel
confirm

DORA (Dataflow-Oriented Robotic Architecture) is middleware designed to streamline and simplify the creation of AI-based robotic applications. It offers low latency, composable, and distributed datafl

Rust Python TOML Markdown C other