
Perspectives on HPC Storage and Interconnects in the Second Half of 2021
$3,500.00
Authors: Mark Nossokoff, Bob Sorensen
Publication Date: 1 202022
Length: 12 pages
Storage and interconnects continue to be important elements of HPC system architecture and are expected to take on even greater significance with increasing demanding and diverse requirements driven by both traditional compute-intensive HPC mod/sim workloads and data-intensive AI workloads. The significance is reflected in recent market data and near-term forecasts, technology adoption and utilization trends, industry announcements in the second half of 2021, and future storage and related technology research direction of Hyperion Research.
Related Products
European Parliament Endorses EuroHPC Exascale Initiative
Steve Conway, Bob Sorensen, Alex Norton, and Earl Joseph
This report looks at the European approach to developing both exascale computers and an indigenous, self-sufficient HPC supply chain in Europe. In recent years the EU and its Member States, along with countries including China, Japan, and the United States, have increasingly recognized the strategic importance of investing more heavily in supercomputing to accelerate scientific and industrial innovation.
July 2018 | Special Report
Perspectives on Composable Systems and HPC/AI Architectures and How They May Fit in the HPC Market
Mark Nossokoff, Bob Sorensen and Earl Joseph
Traditional HPC architectures have typically been designed to address either homogenous workloads (such as physics-based modeling and simulation) with similar, and perhaps more important, fixed, compute, memory, and I/O requirements or, more recently, heterogenous workloads with a diverse range of compute, memory, and I/O requirements. Most HPC data center planners and operators, however, don't have the luxury of focusing on one main type of workload; they typically must support many HPC users and their associated workloads sporting a wide range of compute, memory, and I/O profiles. Architectures have typically consisted of a fixed set of resources, resulting in an underutilized system with expensive elements often sitting idle for long periods of time. One approach being explored to increase system utilization by exposing resources that would otherwise sit idle to appropriately matched jobs waiting in a queue is via composable systems.
May 20 | Special Report