We have provided a reference implementation
on Bitbucket in order to show how the benchmark could be implemented. Please
note that a variety of optimizations should be applied to the code to achieve a
higher performance rate.
- NVIDIA released a container
with a few HPC benchmarks including HPL-AI implementation (requires NVIDIA
Account to download).
- AMD released an optimized
for Zen-4 processors (the page also includes other benchmarks as well: HPL,
HPCG, and STREAM)