Particle.news

Download on the App Store

NVIDIA, OpenAI Set 10 GW AI Buildout With Phased Investment Pledge

The scale concentrates compute supply, drawing scrutiny over cost and access.

Overview

  • OpenAI will deploy at least 10 GW of NVIDIA-based data-center capacity, with the first phase targeted for the second half of 2026 on the Rubin platform, and specifics to be finalized in the coming weeks.
  • NVIDIA said it will invest up to $100 billion in OpenAI in stages tied to each gigawatt deployed, while emphasizing the commitment will not change strategic priorities or product supply for other customers.
  • Antitrust experts warned the arrangement could confer an unfair advantage given NVIDIA’s dominance in GPUs, raising concerns about pricing and delivery terms for rivals.
  • China’s Dongfeng Motor and Huawei unveiled a joint innovation lab focused on in-vehicle software, assisted driving, AI applications and embodied intelligence, followed by a leadership meeting to deepen cooperation across product definition and commercialization.
  • Seres reported cumulative deliveries of AITO M5, M7, M8 and M9 models surpassing 800,000, with the M9 totaling 240,000 and maintaining an 18‑month lead in China’s 500,000‑yuan‑plus segment.