1 briefs tagged #h100.
The benchmark evaluates whether LLMs can optimize GPU kernels on H100 and B200 systems.