alephchang comments

Repositories
Issues
Comments

Results 3 comments of


                                            alephchang

"undefined reference" in building cuDSS examples

Hi Kirill, The default compiler is too old: c++ (GCC) 4.8.5 20150623 (Red Hat 4.8.5-44) I change a new one and add -lstdc++. The undefined reference fixed. Thanks, Yijiang

[QST]how to use one threadblock process one matrix multiplication?

> are you asking for a grouped gather/scatter GEMM? if so, I do not think an example exists today, but it should be easy to build one on top of...

[QST] how to fix the compiling error: static assertion failed with "Vectors implied by the thread map must be divisible by the access type."

I found the compilation error can be fixed by changing the template param AlignmentA/AlignmentB from 8 to 1, any explanation? ``` using GemmKernel = typename cutlass::gemm::kernel::DefaultGemmGrouped< ElementA, LayoutA, cutlass::ComplexTransform::kNone, 1,...