Chipmaker AMD has announced that its record-breaking x86 server GPU was able to surpass the 10 teraflops (FP64) performance barrier, and is expected to be available by the end of 2020.
With parallel processing backed by GPUs, users are able to run more advanced and large-scale programs quicker and more efficiently. AMD’s own Instinct MI100 GPU can now provide up to 11.5 teraflops peak FP64 performance for HPC and up to 46.1 teraflops peak FP32 Matrix performance for AI and machine learning workloads when paired with 2nd-genaration AMD EPYC processors.
“AMD takes a major step forward in the journey toward exascale computing as we unveil the AMD Instinct MI100 – the world’s fastest HPC GPU. Squarely targeted toward the workloads that matter in scientific computing, our latest accelerator, when combined with the AMD ROCm open software platform, is designed to provide scientists and researchers a superior foundation for their work in HPC,” said Brad McCredie, corporate vice president for Data Center GPU and Accelerated Processing at AMD.
The Instinct MI100 is also accelerated by the compute platforms provided by Dell, Gigabyte, HPE, and Supermicro. With its AMD Matrix Core technology built in, the MI100 can also deliver almost 7x boost in FP16 format performance for AI training workloads compared to AMD’s previous generation accelerators.
According to Ravi Pendekanti, senior vice president for PowerEdge Servers at Dell Technologies, their Dell EMC PowerEdge servers are already prepped to support the new AMD Instinct MI100. He noted that customers who need rapid results from HPC and AI workloads will benefit the most with the faster insights from data.
“AMD has been a valued partner in our support for advancing innovation in the data center. The high-performance capabilities of AMD Instinct accelerators are a natural fit for our PowerEdge server AI & HPC portfolio,” Pendekanti said.