1d ago

Developer Cuts Inference Runtime 3x By Fixing GPU Throttling

0
Original post

Solved it. Tracked down the inference + evaluation bottlenecks and ended up cutting runtime by 3x. GPU throttling was absurdly high. Codex genuinely helped a lot and its capabilities are impressive, but after spending hours pushing both Codex and Claude at their limits, I'm even more convinced they're still far from replacing Honestly the satisfaction I got is just another kind of high **Also got crazy number of impressions, it's always the most low effort tweets

3:41 AM · May 18, 2026 View on X