Particle.news

Nvidia Advances Vera CPU Push and Secures Memory Partners After Seoul Talks

The moves are designed to diversify suppliers and build an integrated AI infrastructure platform that reduces bottlenecks and expands Nvidia’s addressable market.

Overview

  • Nvidia held meetings with SK Group and SK Hynix leadership in Seoul on June 7–8 and is expected to announce a cooperation plan with SK following those talks.
  • The company is developing the standalone Vera data‑center CPU and the Vera Rubin AI platform as part of a push beyond GPUs into integrated AI systems.
  • Micron has been certified to supply HBM4 memory for Nvidia’s Vera Rubin platform, a step that joins Samsung and SK Hynix in supporting Nvidia’s next‑generation systems.
  • Nvidia reported very strong fiscal Q1 2027 results driven by data‑center demand and Blackwell adoption, which management says underpins the company’s expansion plans.
  • Nvidia is also deepening ties with neocloud providers such as Nebius and CoreWeave and working with power and packaging suppliers to ensure scale and resilience in large AI factory deployments.