Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


ScaleMP Powers Largest Shared-Memory Systems in Canada

ScaleMP announced that the government of Canada has extended the contract for its large shared memory systems acquired from Dell. These SMP systems use vSMP Foundation to aggregate more than 64 Intel Xeon processors each, totaling more than 1500 CPUs per system. The systems are used for a variety of HPC workloads, including computer-aided engineering (CAE) and computational fluid dynamics (CFD).

Together with our hardware partners, we have been providing technology to the government of Canada since 2012, and are proud of repeatedly earning their business,” said Shai Fultheim, founder and CEO of ScaleMP. “repeat customers are a big part of the vSMP Foundation user community, and we continue to see expansion of our footprint with existing customers along with strong growth in deployments of vSMP Foundation with new ones.”

ScaleMP’s vSMP Foundation delivers a high-end computing solution, aggregating the power of multiple x86 servers into large-scale SMP servers, enabling the creation of the largest scale-up systems in existence. With innovative software-defined computing technology, vSMP Foundation provides a more flexible and cost-effective alternative to legacy, hardware-only SMPs, while increasing overall data center efficiency and reducing operational costs. vSMP Foundation also provides a hyper-converged solution for HPC clusters, with built-in high-performing parallel storage, simplifying cluster installation and operation, and allowing both batch and interactive job execution.

vSMP Foundation aggregates multiple, industry-standard, off-the-shelf x86 servers into a single virtual high-end system. vSMP Foundation provides customers with an alternative to traditional, expensive, symmetrical multiprocessing (SMP) systems, and it also offers simplified clustering infrastructure with a single operating system. It supports aggregation of as many as 128 servers into a single virtual SMP system and enables the transparent use of NVM storage as system memory, providing customers with:

  • Expanded system memory – to as much as 2 PB of shared memory for large-memory requirements – by using DRAM from other servers or by using enterprise-grade non-volatile memory.
  • As many as 32,768 CPUs on a virtual SMP with proven scalability and record-breaking memory bandwidth
    Ease of use and lowest total cost of ownership for clusters coupled with simplified parallel storage architecture
    On-demand SMP with support for many common provisioning systems to enable increased data center utilization with a consolidated architecture.

Sign up for our insideHPC Newsletter

Leave a Comment

*

Resource Links: