The San Diego Supercomputer Center (SDSC) at UC San Diego has announced a partnership with Core Scientific, an infrastructure and software solutions provider for artificial intelligence and blockchain, to offer HPC capabilities to industrial users. SDSC and Core will integrate the Core Scientific Plexus AI software stack with the Expanse petascale supercomputer, launched by SDSC late last year with support from the National Science Foundation (NSF).
Key to Expanse is its ability to support composable systems, the integration of computing elements such as central processing unit (CPU), graphics processing unit (GPU) and other resources into scientific workflows that can include data acquisition and processing, machine learning and traditional simulation. While the main Expanse system is funded by the NSF, SDSC recently installed additional capacity to support industrial workloads.
“Exposing SDSC’s Expanse supercomputer platform via Core Scientific’s Plexus software stack provides users with a consumption-based high-performance computing (HPC) model that not only solves for on-premise infrastructure,” SDSC said in its announcement, “but also has the ability to run HPC workloads in supercomputer centers as well as in the any of the four major public cloud providers—all from a single pane of glass.”
“We are excited to make the power of the Expanse supercomputer available to industry researchers and other customers who need state-of-the-art supercomputer capabilities,” said Ron Hawkins, SDSC director of industry relations. “Core Scientific’s Plexus software stack provides end-customers access to our infrastructure in an intuitive, cloud-like point-and-click experience, which is a potential game changer for supercomputer centers.”
Expanse, which entered service last December, consists of nearly 800 AMD EPYC (Rome) based compute nodes with a 12-petabyte parallel file system and Mellanox InfiniBand HDR interconnect, with more than 100,000 cores available.
“SDSC’s Expanse is a unique and innovative supercomputer that supports a wide range of scientific workloads,” said Core Scientific Chief Product Officer of Artificial Intelligence Ian Ferreira. “Being able to democratize this infrastructure via our Plexus software stack will enable researchers and scientists to seamlessly work between CPU and GPU bound class of problems, all via a single, easy-to-use interface.”
source: SDSC