OpenAI interview: kernel improvements cut GPT-5.6 Sol serving cost by 20%
TheTuringPost · x · 2026-09-16
The Turing Post interviews OpenAI's Philippe Tillet and Matthew Ferrari on using AI to optimize AI: models now find bottlenecks and optimize inference, with kernel improvements cutting end-to-end serving cost of GPT-5.6 Sol by 20%. They also discuss which optimizations waste effort and how well models understand the systems they optimize.
Related event: OpenAI kernel optimizations cut GPT-5.6 Sol serving cost by 20%(2 posts)→
More from Infra
- effect-harness on Cloudflare: run agent loops, tools and credentials on separate VMs — samgoodwin89 · 2026-10-31
- Google's Project Suncatcher boots TPU in orbit and beams back first AI response — sundarpichai · 2026-10-31
- Together AI launches Post-Training platform for agentic RL with tool-use environments — togethercompute · 2026-10-31
- General Compute raises $100M for inference capacity as CoreWeave stock hits 12-month low — YvesMulkers · 2026-10-31
- Analyst warns of unstable AI compute feedback loop as Kioxia's NAND price warning triggers flash crash — bookwormengr · 2026-10-31
- SOMA launches enterprise prompt-compression layer claiming 20%+ inference savings — markjeffrey · 2026-10-31