CardLin/SFEGO_PyOpenCL_Service

This Server.py and Client.py can provide send job to multiple GPU Server to run SFEGO in GPU which can make better throughput.

Spatial Frequency Extraction using Gradient-liked Operator (SFEGO) Service

PyOpenCL Version

Introduction

• This work is based on https://github.com/CardLin/SFEGO_PyOpenCL

• Now, I write a Server and Client that can runs on different GPU Server

Hardware Requirement

• Require GPU on Server to execute OpenCL Kernel Code

• Recommend to use NVIDIA GPU with 1GB+ VRAM (VRAM usage is depend on Image Size)

• AMD Integrated GPU and Intel Integrated GPU can also run this project

• Although It can also run OpenCL on CPU mode but even the Intel Integrated GPU is faster than high-end CPU

Execution

• Modify PLATFORMS = [(0,0,4),(0,1,8)] in Server.py which is (Platform_ID, Device_ID, Worker_Count)

• I have two AMD GPU on this server. [(0,0,4),(0,1,8)] means run 4 thread on (Platform_ID=0, Device_ID=0) and 8 thread on (Platform_ID=0, Device_ID=1)

• python Server.py

• Modify ServerList = [ ("127.0.0.1", 8888, 12), ("192.168.1.33", 8888, 8) ] in Client.py which is ((IP, port, ExecuteCount))

• ExecuteCount is concurrent thread that how many socket connect to specific Server, you can set this number as Worker_Count on server

• Client.py support send image to different server to increase throughput

• Modify IN_Folder for read image and OUT_Folder for save Spatial Frame

• python Client.py