lc0 CPU or GPU if same speed

Hamster · Post by **Hamster** » Mon Jul 27, 2020 12:04 pm

What would you use, the CPU or the GPU if the speed is (almost) the same?

C:\Programs\Engines\lc0\cpu-dnnl>lc0 benchmark
       _
|   _ | |
|_ |_ |_| v0.26.0 built Jul  2 2020
Found pb network file: C:\Programs\Engines\lc0\cpu-dnnl/703350.pb.gz
Creating backend [blas]...
BLAS functions from DNNL version 1.5.0
BLAS max batch size is 256.
>>>>>>>>>>>>>>>>>>>> SNIP >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
===========================
Total time (ms) : 340082
Nodes searched  : 332914
Nodes/second    : 979

and this is the output for the GPU:

Code: Select all

C:\Programs\Engines\lc0\gpu-opencl>lc0 benchmark
       _
|   _ | |
|_ |_ |_| v0.26.0 built Jul  2 2020
Found pb network file: C:\Programs\Engines\lc0\gpu-opencl/703350.pb.gz
Creating backend [opencl]...
OpenCL, maximum batch size set to 16.
Initializing OpenCL.
Detected 1 OpenCL platforms.
Platform version: OpenCL 2.1
Platform profile: FULL_PROFILE
Platform name:    Intel(R) OpenCL
Platform vendor:  Intel(R) Corporation
Device ID:      0
Device name:    Intel(R) UHD Graphics 630
Device type:    GPU
Device vendor:  Intel(R) Corporation
Device driver:  26.20.100.7262
Device speed:   1200 MHZ
Device cores:   24 CU
Device score:   621
Device ID:      1
Device name:    Intel(R) Core(TM) i7-9700 CPU @ 3.00GHz
Device type:    CPU
Device vendor:  Intel(R) Corporation
Device driver:  7.6.0.0814
Device speed:   3000 MHZ
Device cores:   8 CU
Device score:   521
Selected platform: Intel(R) OpenCL
Selected device: Intel(R) UHD Graphics 630
with OpenCL 2.1 capability.
Loaded existing SGEMM tuning for batch size 16.
Wavefront/Warp size: 8

Max workgroup size: 256
Max workgroup dimensions: 256 256 256
>>>>>>>>>>>>>>>>>>>> SNIP >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
===========================
Total time (ms) : 346363
Nodes searched  : 322535
Nodes/second    : 931