MUSA: support ARM64 and enable dp4a .etc (#11843)
* MUSA: support ARM64 and enable __dp4a .etc * fix cross entropy loss op for musa * update * add cc info log for musa * add comment for the MUSA .cc calculation block --------- Co-authored-by: Bodhi Hu <huaishun.hu@mthreads.com>
This commit is contained in:
parent
ee02ad02c5
commit
0b3863ff95
7 changed files with 25 additions and 15 deletions
|
@ -261,6 +261,12 @@ static ggml_cuda_device_info ggml_cuda_init() {
|
|||
GGML_LOG_INFO(" Device %d: %s, %s (0x%x), VMM: %s, Wave Size: %d\n",
|
||||
id, prop.name, prop.gcnArchName, info.devices[id].cc & 0xffff,
|
||||
device_vmm ? "yes" : "no", prop.warpSize);
|
||||
#elif defined(GGML_USE_MUSA)
|
||||
// TODO: refine the .cc to reflect MUSA's actual CC capabilities
|
||||
info.devices[id].smpbo = prop.sharedMemPerBlockOptin;
|
||||
info.devices[id].cc = 100*prop.major + 10*prop.minor;
|
||||
GGML_LOG_INFO(" Device %d: %s, compute capability %d.%d, VMM: %s\n",
|
||||
id, prop.name, prop.major, prop.minor, device_vmm ? "yes" : "no");
|
||||
#else
|
||||
info.devices[id].smpbo = prop.sharedMemPerBlockOptin;
|
||||
info.devices[id].cc = 100*prop.major + 10*prop.minor;
|
||||
|
@ -1782,9 +1788,6 @@ static void ggml_cuda_mul_mat_batched_cublas(ggml_backend_cuda_context & ctx, co
|
|||
}
|
||||
}
|
||||
#else
|
||||
#ifdef GGML_USE_MUSA
|
||||
GGML_ASSERT(false);
|
||||
#else // !GGML_USE_MUSA
|
||||
if (r2 == 1 && r3 == 1 && ggml_is_contiguous_2(src0) && ggml_is_contiguous_2(src1)) {
|
||||
// there is no broadcast and src0, src1 are contiguous across dims 2, 3
|
||||
// use cublasGemmStridedBatchedEx
|
||||
|
@ -1827,7 +1830,6 @@ static void ggml_cuda_mul_mat_batched_cublas(ggml_backend_cuda_context & ctx, co
|
|||
cu_compute_type,
|
||||
CUBLAS_GEMM_DEFAULT_TENSOR_OP));
|
||||
}
|
||||
#endif // GGML_USE_MUSA
|
||||
#endif
|
||||
|
||||
if (dst->op_params[0] == GGML_PREC_DEFAULT) {
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue