NVIDIA has launched the Vera Rubin NVL72, which is ramping up production across partners including CoreWeave, Google Cloud, Microsoft Azure, and Oracle Cloud Infrastructure. This platform boasts the largest rack-scale supply chain, spanning over 350 factory sites in 30 countries, designed to meet the increasing compute demands of AI applications.
The Vera Rubin platform is engineered for optimal performance, delivering the highest performance per watt and the lowest token cost. CoreWeave's benchmark indicates a tenfold increase in throughput per megawatt compared to the Grace Blackwell NVL72, highlighting its efficiency for power-constrained AI factories. The system integrates seven chips and five rack trays, including the NVIDIA Vera CPU, which enhances single-threaded performance and reduces memory latency significantly.
Looking ahead, NVIDIA's innovations in networking and cooling technologies, such as the sixth-generation NVLink and Spectrum-X Ethernet, promise to further enhance AI infrastructure capabilities. No further timeline was disclosed at the time of publication.
Editor's Note
NVIDIA's Vera Rubin NVL72 represents a significant advancement in AI infrastructure, particularly for enterprises focused on maximizing efficiency and performance. The integration of advanced cooling and networking technologies positions it as a competitive solution in the rapidly evolving landscape of AI and cloud computing.
Leave a comment