A regular task we all do is check the weather daily to know what to wear over the few days ahead, especially looking forward to sunny days in the upcoming winter. In years past, the meteorologist’s tools of future prediction were limited to atmospheric data from a key source like the National Oceanic and Atmospheric Administration (NOAA) for the region, while local predictions were not as comprehensive.
Today’s weather forecasters are able to fully define and communicate a local forecast real-time, in minute-by-minute detail. High performance computing (HPC) has led innovation such that today’s solutions rapidly turn real-time data into visual insights to predict a warm winter with more sun or understand factors of winter storm weather which your local weather forecaster can use to forewarn their audience.
The rapid pace of technology and the ability to deliver more innovation from the traditionally specialized infrastructure has become more pervasive in enterprises, data centers and businesses of all sizes as well as academia, financial services and life sciences. Furthermore, the traditional server and central processing unit (CPU) which drives many standard workloads in business has found a way to serve more HPC-type workloads by offloading to graphics processing units (GPUs).
This approach not only builds upon standard infrastructures which can be virtualized and scaled into clusters running the HPC workloads, they can also be managed and deliver secure operations, without the need for additional tools as specialized infrastructures require. Dell EMC PowerEdge servers deliver the ability to drive HPC workloads with the powerful AMD Instinct™ MI100 accelerators, introduced last year and supported on the AMD EPYC™ powered PowerEdge R7525 server and newer R750xa platforms.
These powerful platforms are ideal for driving complex HPC workloads such as molecular dynamics, physics and other clustered applications on PowerEdge servers, along with the AMD CDNA™ architecture, a compute optimized GPU architecture to deliver high performance, in the MI100 and AMD’s ROCm™, the first open-source exascale-class platform for accelerated computing.
Now, with the faster pace of innovation supporting the demand for faster visualization and outcomes, organizations have more capabilities to drive HPC projects with the newest members of the GPU family, the AMD Instinct™ MI200 series accelerators, and the announcement of the AMD Instinct™ MI210 accelerator, with planned support on Dell PowerEdge servers.
As HPC continues to push the leading edge of discovery and outcomes in artificial intelligence (AI) and machine learning (ML) to simulation and other complex workloads, the infrastructure supporting HPC must also keep pace while enabling businesses to start new projects of any size and with a framework and ecosystem of partners to provide tools and guidance along the way. Dell Technologies enables customers with a broad range of strategy and solutions such as:
Creating new value in HPC requires an infrastructure that grows with, and from, innovation. With Dell PowerEdge servers, the next generation of purpose-built GPUs like the AMD Instinct™ MI210 GPU, Dell Validated Designs and Dell Centers of Excellence, HPC helps research institutions to enterprises extract more value from their data, accelerating discovery and analysis of current weather – sunny skies versus winter storms – in order to predict impacts sooner and drive actionable decisions faster.
Learn more about accelerators for Dell EMC PowerEdge servers here.
Join us at SC21, running November 14-19, 2021.