Spconv support gpu arch

[xiaoxifuhongse]

gpu : NVIDIA GeForce RTX 2080 SUPER system: ubuntu20.04 cuda: 11.7

when i run Lidar_AI_Solution-master/Libraries/3DSparseConvolution/tool/pytest.py then CUDA Runtime error spconv::setup_hash_and_route<<<(numinput+1023)/1024,1024,0,stream>>>(hash.get(),routemask.ptr(),route.ptr(),indices,num_input,prob.kv).# no kerenl image is availabel for exectuion on the device, code=cudaErrorNoKernelImageForDevice[209]in file src/spconv/rulebook.cu:150 Turing, or Ampere gpu how to handle it

[lix19937]

RTX 2080 sm is 75, not supported by libspconv.so which need sm>=80, so it supports Ampere, like RTX3070, 3080, or ORIN.

[yunqiushuiman]

Thank you for the expert's answer to this question. I am facing the same issue as well. CUDA Runtime error init_quantization_scale<<<grid, block, 0, _stream>>>( input_scale, outputscale, weight, prodDHWI, int8weight.ptr(), weightscales.ptr(), alpha_.ptr(), (half *)bias, biasscaled.ptr(), numel) # no kernel image is available for execution on the device, code = cudaErrorNoKernelImageForDevice [ 209 ] in file src/spconv/implicit-gemm.cu:241 tool/run.sh: line 41: 11487 Aborted (core dumped) ./build/bevfusion $DEBUG_DATA $DEBUG_MODEL $DEBUG_PRECISIONIf I want to run this program on a low-powered laptop, what can I do to make it work? How can I modify the program to make it suitable for a computing power of 75?

[yunqiushuiman]

I want to run this program on a low-powered laptop, what can I do to make it work? How can I modify the program to make it suitable for a computing power of 75?

[hopef]

Sorry. Currently, the libspconv.so can not run on sm75 or lower.