NVIDIA Unveils Vera Rubin Platform With 7 Exaflops for Scientific Supercomputing
Updated
Updated · GlobeNewswire · Jun 22
NVIDIA Unveils Vera Rubin Platform With 7 Exaflops for Scientific Supercomputing
3 articles · Updated · GlobeNewswire · Jun 22
Summary
NVIDIA said Vera Rubin can deliver more than 7 exaflops of AI-for-science performance and 5 petaflops of native FP64 compute, targeting climate modeling, fluid dynamics, quantum chemistry and energy exploration.
The rack-scale platform combines Rubin GPUs, Vera CPUs and NVIDIA networking in a direct liquid-cooled design, letting one system run simulation, AI training, inference and real-time data analytics together.
Three major labs have already lined up systems: LRZ’s Blue Lion is slated for 2027 with about 30 times its current computing power, while Berkeley Lab’s Doudna and Los Alamos’ Mission, Vision and Veritas will use Vera Rubin-based designs.
Bull, Dell, GIGABYTE, HPE and Supermicro are preparing custom high-density racks with up to 144 GPUs, with Vera Rubin NVL4 systems expected to reach the market in the fourth quarter.
Will NVIDIA’s new supercomputer, designed for 'agentic AI,' fundamentally alter the scientific discovery process itself?
As NVIDIA builds a fully integrated 'AI factory,' can any competitor realistically challenge its overwhelming market dominance?
NVIDIA Vera Rubin Platform Unveiled: A Paradigm Shift in High-Performance Computing and Autonomous AI Workloads
Overview
The NVIDIA Vera Rubin platform, unveiled in June 2026, marks a major step forward in scientific computing by unifying high-performance computing and artificial intelligence. It is designed to handle demanding scientific workloads by providing native FP64 capabilities for highly accurate simulations and delivering robust AI performance for advanced model development. This integration means researchers no longer need separate systems for simulations and AI tasks, allowing them to work more efficiently. By combining these strengths, the Vera Rubin platform enables scientists to accelerate discovery and innovation across a wide range of research fields.