gpu-unlock boosts RTX 3090 Ti BF16 GEMM 87.5→157.4 TFLOPS
jiqizhixin · x · 2026-10-26
The open-source gpu-unlock project uses a Linux kernel module to lift FP16/BF16 instruction-rate limits on Ampere RTX 30-series cards without patching NVIDIA drivers, combining the CMP unlocker's V67/PLM technique with pre-driver UEFI timing. On an RTX 3090 Ti, BF16 GEMM jumps from 87.5 to 157.4 TFLOPS (80% more), with reported LLM prefill gains of 14–36%. The repo has 7.1k stars, and Claude is listed as the top contributor on GitHub.
Related event: Enthusiasts Unlock Hidden Tensor Power on RTX GPUs(2 posts)→
More from Infra
- Compute rental rates rise across the curve as AI buildout funding tightens, Meta 10-Q shows — AccBalanced · 2026-10-26
- Open-source recipe: abliterated DeepSeek-V4.1-Flash at ~200 tok/s on 4x RTX PRO 6000 — burny_tech · 2026-10-26
- Micron opens second advanced packaging plant in Houli, a concrete HBM step — tengyanAI · 2026-10-26
- What do you do with the spare 8 GPUs in a 64-card NVL72 setup? Keep them hot as spares — _xjdr · 2026-10-26
- Jalapeño chip is just a systolic array with HBM: why custom inference silicon is easier than you think — ai · 2026-10-26
- Torch.compile Flag Cut 16 Regions to 2, Delivering 4.53x Speedup on B200 — cfregly · 2026-10-26