Nvidia 1: NVIDIA Selene: Leadership-Class Supercomputing Infrastructure
DescriptionAI and HPC continue to expand the frontiers of exploration, but today’s most challenging opportunities require a new architecture that can deliver unprecedented levels of computing scale to drive faster insights from massive volumes of data. Where traditional high-performance compute clusters take months or years to deploy, NVIDIA’s Selene supercomputer, built on the NVIDIA DGX SuperPOD™ reference architecture, offers a faster path to massive scale, in a Top10-class supercomputer that was deployed in under a month. Join us for this special insider’s look at how Selene and the DGX SuperPOD are redefining AI infrastructure. We’ll explore what fueled Selene’s design, how the design can be deployed as a commercially available turnkey solution, how we rapidly deployed Selene using its modular design and ability to incrementally scale, and the tools we use to run mission-critical mixed workloads.