Los Alamos National Laboratory is preparing new HPE and NVIDIA supercomputers that use NVIDIA Vera CPUs to support scientific discovery.
The announcement is notable because it connects next-generation CPU architecture with agentic scientific AI, where systems may help reason across simulations, experiments, and data analysis.
That kind of workload needs more than accelerators alone. Memory, networking, and CPU design all shape how effectively researchers can run complex AI-assisted science pipelines.