Benchmark AMD MI100 GPUs with Koi Computers Demo Cluster

By Jeanna Van Rensselar | 22nd Mar 2021

We are an AMD Elite Partner for turnkey HPC servers and clusters and are offering a free trial of AMD’s groundbreaking MI100 accelerator! The MI100 is actually the world’s fastest HPC GPU!

Powered by the first AMD CDNA™ architecture, the MI100 accelerators deliver a giant leap in compute and interconnect performance, offering 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. Features and benefits include:

  • All-new AMD CDNA architecture delivering as much as 74% higher FP64 performance to accelerate HPC workloads when compared with AMD prior gen*
  • All-new Matrix Core Technology delivering a giant leap in AI matrix performance with FP32 and FP16--introducing new bFloat16 operations
  • 2nd Generation Infinity Architecture bringing 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**
  • Enhanced RAS and security capabilities
  • AMD ROCm™ open software ecosystem

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 newest 3rd Gen AMD EPYC™ Milan processors and AMD Infinity Fabric™ technology, MI100-powered systems provide scientists and researchers with extremely powerful platforms.

At Koi Computers, we fully leverage the AMD MI100 with expertly engineered servers and clusters that are completely customized to accommodate your current and projected needs. To sign up for a demo or learn more visit: https://koicomputers.com/amd-mi100/