NVIDIA Launches Vera Rubin Supercomputing Platform for AI & Science
NVIDIA announced the launch of its Vera Rubin platform, designed to accelerate the world's most demanding HPC workloads and bring agentic AI to scientific computing. The platform combines NVIDIA Rubin GPUs and Vera CPUs, offering over 7 exaflops of AI for science and 5 petaflops of native FP64 support, with up to 144 GPUs per rack. Leading supercomputing centers, including Leibniz Supercomputing Centre, National Energy Research Scientific Computing Center, and Los Alamos National Laboratory, are adopting Vera Rubin for their next-generation systems. Global system manufacturers are also preparing to offer custom high-density Vera Rubin supercomputing systems, with availability expected in Q4 this year.
Key Highlights
- NVIDIA launched the Vera Rubin platform for scientific supercomputing, combining Rubin GPUs and Vera CPUs.
- The platform delivers over 7 exaflops of AI for science and 5 petaflops of native FP64 support.
- Vera Rubin systems can support up to 144 GPUs per rack for extreme memory bandwidth.
- Leibniz Supercomputing Centre, NERSC, and Los Alamos National Laboratory are adopting Vera Rubin for next-gen supercomputers.
- Global system manufacturers like Bull, Dell, GIGABYTE, HPE, and Supermicro will offer Vera Rubin systems.
- Vera Rubin NVL4-based systems are expected to be available from manufacturers in Q4 this year.
Price Impact
More from NVDA