RTX 30-series GPU Unlock tested: up to 59% throughput gain in ComfyUI on 3090 Ti
under-_-null · reddit · 2026-10-24
A Reddit user benchmarked Zagitta/gpu-unlock, which raises the FP/BF16 instruction issue rate of consumer RTX 30-series cards to professional-card levels, on a 3090 Ti running ComfyUI — with 47–59% throughput gains across BF16 models.
Key results (median time per image):
- MOSS-Image SFT BF16: 71.98s → 45.40s (+58.6%)
- Qwen-Image 2.1 Turbo BF16: 8.51s → 5.41s (+57.2%)
- Z-Image Turbo BF16: 6.79s → 4.35s (+56.1%)
- FLUX.1 dev FP8 checkpoint: 31.96s → 20.66s (+54.6%)
- INT8 quantized models saw almost no benefit (+0.3% to +4.6%)
Methodology: 15 images per config after 3 warmups, same prompts/seeds, and all 120 stock/unlocked image pairs decoded to identical RGB pixels — lossless output. Setup: Ubuntu, NVIDIA 615.71.09, ComfyUI 0.42.1, batch size 1, stock 450W power limit. Similar gains expected on other supported 30-series cards.
More from Infra
- Video DiT runs on 12GB VRAM at same speed via int8, model weighs only 9GB — cocktailpeanut · 2026-10-24
- Modal and Sakana AI optimize Namazu API to the kernel: NVFP4 quantization, DFlash drafting, KV routing — blelbach · 2026-10-24
- Semis debate: is high-NA EUV (EXE:5200) cost-effective for 18A nodes as TSMC demands tool upgrades — pstAsiatech · 2026-10-24
- Intel B60/B65/B70 GPU prices jump 30% in a day, B70 heading to $1,840 — SamSausages · 2026-10-24
- RTX 3070 runs quantized Qwen3.8 at 1000 tps prefill with Strata — full config shared — SetNo7692 · 2026-10-24
- Cloudflare cuts CPU use 30% on D1 backend with new profiler, Workers Observability drops memory/CPU ~50% — irvinebroque · 2026-10-24