OpenAI and Broadcom unveiled Jalapeño, OpenAI's first custom AI chip, designed to serve the company's inference workloads more efficiently and at lower cost than existing alternatives. Announced on Wednesday, the chip is planned for initial deployment by end of 2026, with significant scaling through 2027 and into early 2028. This move reflects OpenAI's strategy to build its own computing stack and address an acute shortage of AI processing capacity as demand from its customers grows beyond what existing suppliers can provide.
Summaries like this, in your inbox every morning.
Sign up free →What happened
OpenAI and Broadcom on Wednesday announced Jalapeño, OpenAI's debut custom silicon chip designed to run AI models in production. The companies designed the chip in nine months and plan initial deployment by the end of 2026, with scaling through 2027 and into the first half of 2028.
Why it matters
OpenAI says it 'cannot get compute fast enough' and Broadcom's CEO confirmed that demand from his six largest customers is 'simply insatiable.' By designing its own chips alongside the broader computing stack, OpenAI aims to serve more AI capability with greater efficiency, reducing its dependence on Nvidia's expensive GPUs and addressing a critical supply constraint.
What to watch
Jalapeño is an ASIC (application-specific integrated circuit), which industry experts say is less flexible than Nvidia's GPU but also less expensive and customizable for specific AI tasks. The companies ultimately aim to build enough capacity to require 10 gigawatts of power, signaling massive infrastructure growth over the coming years.
No comments yet. Be the first to share your thoughts!
Log in to join the discussion
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
5 minutes a day. The AI essentials.
200+ sources · Email / LINE / Slack
