AMD MI100 Demo on Koi Computers HPC Cluster

By Jeanna Van Rensselar | 22nd Mar 2021

As an AMD Elite Partner for ready-to-deploy servers and clusters, we are one of the leading turnkey HPC technology providers in the U.S We are now offering a free demo of AMD’s lightening fast MI100 GPU—the fastest HPC GPU in the world!

The MI100 is powered by the first AMD CDNA™ architecture. When compared to AMD’s earlier accelerators, the MI100 features a nearly 3.5x boost for HPC (FP32 Matrix) performance and a nearly 7x boost for AI throughput (FP16) performance compared to AMD’s prior generation accelerators. Highlights of this exciting GPU include:

  • Matrix Core Technology creates a major boost in AI matrix performance with FP32 and FP16 and new bFloat16 operations
  • AMD CDNA architecture delivers up to 74% better FP64 performance to accelerate HPC workloads.
  • 2nd Generation Infinity Architecture enables advanced platform connectivity and scalability
  • Enhanced HBM2 memory interface with about 20% more theoretical memory throughput and improved latencies vs. AMD’s prior gen accelerators.
  • AMD ROCm™ open software ecosystem
  • Enhanced RAS and security capabilities

MI100 accelerators offer an open environment that does not lock users into a vendor-specific solution, enabling developers to enhance existing GPU codes to run everywhere. Combined with the powerful 3rd Gen AMD EPYC™ processors and AMD Infinity Fabric™ technology, MI100-powered systems deliver extremely powerful platforms.

The MI100 is just the latest in AMD’s spectacular run of powerful products that bring new capabilities to HPC. These products fill gaps in the HPC market and create game-changing opportunities for researchers, engineers and everyone that counts on HPC for fast, reliable results. To learn more visit: https://koicomputers.com/amd-mi100/