hopper_cublaslt_bf16_check
Smaller Hopper cuBLASLt BF16 correctness lane with reference enabled.
This repository contains the standalone CUDA source for the hopper_cublaslt_bf16_check lane from
the PyC kernel lab. It is a source artifact for inspection and benchmarking;
it is not a precompiled binary and the result below is not a universal ranking.
Performance
| Kernel | GPU / architecture | Shape | Best recorded result | Evidence |
|---|---|---|---|---|
hopper_cublaslt_bf16_check |
not recorded | not recorded | Not measured in the published campaign | No published performance receipt was found for this lane. |
The result is reported with the original campaign's timing and correctness context. Compare kernels only when GPU, CUDA version, matrix shape, warmup, repeats, and reference/correctness mode match.
Source
kernel.cu— copied fromkernels/prototypes/hopper/cublaslt_bf16/kernel.cu.- Original lane tags:
cuda, matmul, hopper, sm90, prototype, cublaslt, bf16, correctness.
Build/run contract
{nvcc} -O3 -std=c++17 -lineinfo -gencode arch=compute_90,code=sm_90 -gencode arch=compute_90,code=compute_90 {source} -lcublasLt -lcublas -o {build_dir}/{name}
{build_dir}/{name} 512 512 512 2 10 0
- Downloads last month
- -
cuda
kernel
gpu-optimization
hpc
apache-2.0