Nvidia (Nasdaq: NVDA) has rolled out a central processing unit designed to help customers manage high-performance computing and artificial intelligence workloads at data centers.
Grace CPU uses processor technology from U.K.-based semiconductor vendor Arm and supports natural language processing model training, Nvidia said Monday.
Jensen Huang, founder and CEO of Nvidia, said the company licensed and applied Arm intellectual property to develop the Grace system.
Huang added the product offers what he calls a “third foundational” computing technology and seeks to re-architect data center operations.
The U.S. Department of Energy’s Los Alamos National Laboratory and Switzerland’s national supercomputing center plan to adopt the new CPU technology for integration into supercomputing machines Hewlett Packard Enterprise (NYSE: HPE) will build for scientific research work at the two facilities.