OpenAI's Astra migrates to Vera Rubin, doubling inference throughput via self-optimization

OpenAI migrated its Blackwell-optimized Astra code to NVIDIA's new Vera Rubin platform, gaining 3x throughput out of the box, then another 2x after 72 hours of recursive self-optimization shared at AI Infra Summit.

2026-09-16 ~ 2026-09-17 · 2 related posts

• Astra optimizes its own inference on Rubin chips, doubling throughput in 72 hours — bookwormengr · 2026-09-16
• OpenAI's Astra gets 3x throughput on NVIDIA Vera Rubin, plus 2x more in 72 hours — MickeySteamboat · 2026-09-17