ChipsAug 30, 2026 · 15:374 sources

OpenAI Jalapeño Outperforms Nvidia GB300 With AI Code Humans Cannot Read

From huggingnews.com

Image: huggingnews.com
AI;DR

OpenAI's Jalapeño chip, developed with Broadcom, delivered higher tokens per watt and lower latency than Nvidia's GB200 and GB300 on specific models. The hardware utilizes AI-generated assembly and kernels written in Gluon, a language OpenAI engineers cannot reason about line by line. Initial deployments are planned before year end, with full scale operation starting in 2027.

Read the sourceSign in to suggest

Sources

+ 2 more
  1. SUPPORT@jukan05“AI may even begin designing architectures that humans would find awkward, unintuitive, or outright unpleasant to work with”x.com
  2. SUPPORT@beth_kindig“initial deployments expected before year end”x.com

One ranked edition, every morning

The stories that matter today. Free, cancel anytime.

Telegram @aidr_today