Key Highlights

  • Vera Rubin platform delivers over 7 exaflops of AI performance and 5 petaflops of native FP64 in a single rack, with systems from Dell, HPE, and Supermicro due in Q4 2026.
  • NVIDIA Halos for Robotics is the first full-stack safety system for physical AI, with Agility's humanoid robot Digit among the first to integrate the architecture.
  • A record 35 AI supercomputers are now in development across 23 European countries, backed by more than 800 AI exaflops deployed or announced since last year.
  • NVDA closed at $208.65 on June 22, with a market cap of approximately $5.05 trillion, a P/E ratio of around 32x, and EPS of $6.53.

NVIDIA's Vera Rubin platform, unveiled at ISC High Performance 2026 in Hamburg, is designed to combine simulation, AI training, and data analytics in a single rack-scale system. The platform supports up to 144 GPUs per rack and delivers performance on par with systems on the TOP500 list of the world's most powerful supercomputers. Leibniz Supercomputing Centre, the National Energy Research Scientific Computing Center, and Los Alamos National Laboratory have all selected Vera Rubin for their next-generation systems, targeting workloads spanning climate modeling, nuclear energy, and national security. Systems are expected to be available from global manufacturers in Q4 2026.

Separately, NVIDIA announced Halos for Robotics, positioning it as the first unified safety architecture for physical AI systems. Built on more than 18,600 engineering years of autonomous vehicle safety development, the system integrates AI compute, sensor connectivity, safety software, and a newly accredited inspection lab. Agility, whose Digit humanoid robot operates in logistics and warehouse facilities for clients including Amazon and Toyota Motor Manufacturing Canada, is the first company to incorporate Halos for Robotics into its safety stack. The NVIDIA Halos AI Systems Inspection Lab has received ANAB accreditation, providing a recognized pathway for third-party certification.

On the infrastructure front, NVIDIA confirmed that a record 35 AI supercomputers are under development across Europe, with systems spanning 23 countries and serving more than 3 million researchers. More than 90% of Europe's AI factory buildout is running on NVIDIA infrastructure, with 800 AI exaflops deployed or announced in the past year. Projects include Barcelona Supercomputing Center's MareNostrum5 AI expansion, Germany's first AI factory at HLRS, and Italy's IT4LIA system with over 8,000 GPUs.

For investors tracking NVDA stock, the announcements reinforce the company's expanding role across scientific computing, industrial robotics, and sovereign AI infrastructure, markets that extend well beyond its core data center GPU business. NVDA traded at approximately $202.53 in pre-market activity on June 23, reflecting broader tech sector pressure rather than company-specific concerns.

 

This article is for informational purposes only and does not constitute financial advice. Please consult a licensed financial adviser before making investment decisions.