RTX 4090 volatile unlock boosts FLUX.1-dev inference 64% to 214 TFLOPS
BrightRaiderDev · reddit · 2026-11-02
A developer published reproducible benchmarks and an open-source "universal unlock" for the RTX 4090 (Ada Lovelace) that raises FLUX.1-dev throughput by 64.2% under Windows.
Key results
- FLUX.1-dev (ComfyUI, fp8, 1024x1024, 20 steps): 34.2s → 24.1s (1.20 it/s → 1.97 it/s).
- FP16 GEMM (4096x4096): 132.1 → 213.79 TFLOPS (+61.8%); FP32 unchanged at 82.5 TFLOPS, confirming standard CUDA pipelines stay intact.
- Endurance test: 20 consecutive 1280x1280 generations over 12m36s, 23.5 GB VRAM peak, max 66°C core / 75.8°C hotspot, 342.4 W average power, flatline 2,759 MHz, zero throttling, 20/20 bit-perfect images with no CUDA errors.
Method
A transient, volatile pre-boot hardware state reconfiguration via UEFI before the Windows display driver initializes — the GPU EEPROM/VBIOS is untouched and write-protected, and everything reverts to stock on reboot. Full docs and implementation on GitHub (ada-4090-universal-unlock).
Related event: Community Fix Unlocks NVIDIA's Windows Tensor Throttling on RTX 40/30 GPUs(2 posts)→
More from Infra
- AI Defense Agents Running 24/7 Could Send Token Costs Through the Roof — robleclerc · 2026-11-02
- LookAtMyQuant: a quantization compiler with 12x lower KL at Q4_K_M size on Gemma 4 E2B — teortaxesTex · 2026-11-02
- AI Data Centers Near a PC-Style Inflection Point: OCP Analysis of Power, Cooling, Networking — ai · 2026-11-02
- Cerebras halves in 5 months post-IPO: $25.4B order book leans heavily on OpenAI — FinanceYF5 · 2026-11-02
- Together launches post-training platform aligning RL training with serving — togethercompute · 2026-11-02
- Demis Hassabis: AI data centers should be beautiful — Recent_Fox4339 · 2026-11-02