Particle.news
Download on the App Store
154 ARTICLES
·
2h ago

OpenAI’s Broadcom Chip Deal Escalates AI Arms Race as Oracle Commits 50,000 AMD GPUs

The buildout shifts to delivery in 2026, with gigawatt-scale projects drawing fresh scrutiny over financing and power.

AI (Artificial Intelligence) letters and robot hand are placed on computer motherboard in this illustration created on June 23, 2023. REUTERS/Dado Ruvic/Illustration/File Photo
A smartphone with a displayed Broadcom logo is placed on a computer motherboard in this illustration taken March 6, 2023. REUTERS/Dado Ruvic/Illustration
Larry Fink AI
Image

Overview

  • OpenAI and Broadcom will co-develop and deploy 10 gigawatts of custom accelerators from 2026 through 2029, a disclosure that helped add more than $150 billion to Broadcom’s market value.
  • Oracle Cloud Infrastructure said it will roll out 50,000 AMD Instinct MI450 GPUs starting in the second half of 2026, expanding AMD’s foothold as an alternative to Nvidia for AI workloads.
  • Nvidia previously announced a pact to provide at least 10 gigawatts of systems to OpenAI and intends to invest up to $100 billion progressively, with the first gigawatt slated for the second half of 2026.
  • Cumulative OpenAI-linked infrastructure commitments are now reported at roughly $1 trillion, including a confirmed $300 billion Oracle agreement, intensifying questions about concentrated supply and circular financing.
  • Analysts are split between bubble-risk warnings and arguments that demand remains robust, while execution risks loom over multi-year deployments and the massive power required for gigawatt-scale data centers.