85.3 GFlops: Optimizing FP32 Matrix Multiplication on a Single AMD Zen 3 Core

Part of

[HuggingFace] Add IntelligenceLab/Long-Horizon-Terminal-Bench to the Benchmark allow-list