houslast3/How-I-Squeezed-31.9-TFLOPS-out-of-an-RTX-4060-Beating-cuBLAS-by-14-

2 best points · 1 comments · 1 Hacker News post · 1 GitHub stars · View on GitHub

From the README

CUTLASS/WMMA GEMM Benchmarks — NVIDIA RTX 4060 (Ada Lovelace) Systematic benchmark of FP16→FP32 GEMM optimization strategies on a consumer-grade Ada Lovelace GPU, comparing cuBLAS, CUTLASS device-level API (custom pipeline depths and tile sizes), and a hand-tuned WMMA Tensor Core kernel.

Hacker News discussions

• I Squeezed 31.9 TFLOPS out of an RTX 4060, Beating cuBLAS by 14% (2 points, 1 comments, by houslast, 2026-07-13)

More from houslast3 on Hacker News

• houslast3/85.30-GFLOPS-Single-Core-FP32-Matrix-Multiplication-on-AMD-Zen-3 (93 points): 85.3 GFlops: Optimizing FP32 Matrix Multiplication on a Single AMD Zen 3 Core