NVIDIA Reveals Rubin HPC GPU With NVLink 7 And Over 50,000 CUDA Cores
by
Paul Lilly
—
Tuesday, November 17, 2026, 09:40 AM EDT
Rubin HPC is a new GPU in the NVIDIA Vera Rubin platform - Image: NVIDIA
NVIDIA is cramming a whole lot of CUDA cores into its next-generation Rubin HPC GPU, a new entry to its Vera Rubin platform. Announced at the SC26 supercomputing conference, NVIDIA's Rubin HPC GPU is a monster chip with more than 50,000 CUDA cores on a single processor. That's a record for NVIDIA, though sheer core count is only part of the story.
The other notable highlight is that NVIDIA's newest accelerator introduces seventh-generation NVLink connectivity to keep all those cores fed and busy crunching heaps of data. Beyond some previous roadmap teasers, NVIDIA has not disclosed many details about NVLink 7, but as a point of context, NVLink 6 enables 3TB/s of bandwidth per GPU.
While NVIDIA's Rubin HPC joins the firm's broader Vera Rubin platform, it's tailored for high-performance computing (HPC) rather than focusing primarily on training and running AI models. Think complex scientific simulations, engineering calculations, and other workloads where moving massive datasets efficiently can be just as important as processing them.
⚡ Faster simulation means faster discovery.
Introducing NVIDIA Rubin HPC, a new GPU in the NVIDIA Vera Rubin platform designed to accelerate simulation and data processing for science, engineering and analytics workflows.
Delivering 3x the performance of NVIDIA Blackwell and with more than 50,000 CUDA cores, it accelerates the path from exploration to scientific discovery and industrial design. #SC26
🔗 https://t.co/vGwOcSlrzd
Of course, there is an AI angle here, too. NVIDIA envisions a future in which AI agents help scientists and engineers generate and evaluate new ideas, relying on GPU-powered simulations to determine which ones are worth pursuing.
"For agents tackling science and engineering, faster simulation means faster discovery," said Ian Buck, vice president of hyperscale and high-performance computing at NVIDIA. "Rubin HPC accelerates the simulations agents use to test ideas and guide their next steps, helping scientists and engineers explore more possibilities and bring breakthroughs closer."
All that said, it's the number of CUDA cores that really makes Rubin HPC stand out. To put it into perspective, NVIDIA's standard Vera Rubin GPU packs 28,672 CUDA cores and 224 streaming multiprocessors, which itself dwarfs the consumer-grade GeForce RTX 5090 with its 21,760 CUDA cores.
Beyond the massive number of CUDA cores and NVLink 7 connectivity, NVIDIA has yet to reveal several important specifications, including memory capacity, bandwidth, and power consumption. Those details will come at a later date.
As for availability, NVIDIA says its Rubin HPC GPU is expected to arrive in 2028.
Paul is a seasoned geek who cut this teeth on the Commodore 64. When he's not geeking out to tech, he's out riding his Harley and collecting stray cats.
TOP STORIES
Which New GPU Is For You?
KEEP INFORMED
Stay updated with the latest news and updates. Subscribe to our newsletter!