RTX 3090 power-limit unlock doubles inference speed from 75T to 135T
AIFlow_ML · x · 2026-10-25
Developer KBlueleaf confirmed a successful RTX 3090 unlock: raising the power wall to 360W nearly doubled measured PyTorch inference throughput from 75T to 135T. The mod uses a specific commit (a9b0c75) of a community tool. Commenters joked that 'nobody will be able to buy a GPU anymore' as old cards become far more cost-effective.
Related event: KohakUnlock Unlocks Hidden Compute Power on RTX GPUs(3 posts)→
More from Infra
- Running big models on a single RTX 4090: INT8 weights, FP8 decoder, RAM tiering — AIFlow_ML · 2026-10-25
- ByteDance, Alibaba and Tencent order 3M+ Nvidia RTX PRO 5500 cards worth $35B+ — Miles_Brundage · 2026-10-25
- Nebius hits 67B tokens/day on OpenRouter, but rivals process up to 2T daily — AccBalanced · 2026-10-25
- OpenRouter analysis finds evidence of spoofing-style price manipulation by inference providers — AccBalanced · 2026-10-25
- EmbeddedLLM receives AMD MI350X server donation to advance vLLM ROCm work — AccBalanced · 2026-10-25
- 106 Tesla Megapacks to replace a former Belgian coal plant with 400 MWh storage — XFreeze · 2026-10-25