The Center for Computational science at The University of Tsukuba Creates Pegasus, a ‘Huge Memory’ AI Supercomputer with The Supermicro Superblade®

The Center for Computational Sciences at the University of Tsukuba is a multidisciplinary center that partners with many organizations on basic research in several areas. As part of an ongoing project to deliver state-of-the-art computing capabilities to a wide range of researchers, the University of Tsukuba, with NEC (as the prime contractor), has created a supercomputer that will address the computing and storage needs using Supermicro SuperBlade. The system is among the first in the world to use NVIDIA H100 GPUs and 4 th Gen Intel® Xeon® Scalable processors.

Challenges The Center for Computational Sciences at the University of Tsukuba determined that a new HPC system was needed to address the increasing needs of researchers. As new AI research, Data Science, and Computational Science applications are used and developed, a system with significant amounts of memory per process is needed to address these initiatives. While on the contrary, the technical trend is that the number of cores per CPU continues to increase, and the amount of RAM per core continues to decrease. The University of Tsukuba needed a solution that addressed a number of requirements: fast CPU processing, extensive memory access, and the latest GPU processing.

Solution The Center for Computational Sciences at the University of Tsukuba selected the Supermicro SuperBlade® for its new supercomputer that addresses large memory needs with the latest GPU technologies for HPC and AI applications. Specifically, the University of Tsukuba acquired 120 Supermicro 6U SuperBlades (in 24 enclosures, 5 SuperBlades per enclosure). Each SuperBlade (SBI-611E-5T2N) blade server is powered by 4th Gen Intel® Xeon® Platinum 8468 processors (350W TDP) and contains a single NVIDIA® H100 Tensor Core GPU 80GB HBM2E and 128GB DDR5- 4800MHz of memory per blade. In addition, each node contains Intel® Optane™ Persistent Memory 300 series. Utilizing the NVIDIA Quantum-2 400Gb/s InfiniBand networking platform, each blade uses the NVIDIA ConnectX®-7 HCA to connect to the NVIDIA Quantum-2 switch. Below is an image of SuperBlade used in the University of Tsukuba’s new

supercomputer Pegasus.

Shopping cart


No products in the cart.

Continue Shopping