Hewlett Packard Enterprise fuels innovation with enhanced end-to-end high performance computing (HPC) portfolio

November 13, 2018 • Blog Post • Bill Mannel, Vice President and General Manager, HPC and AI


  • HPE showcases latest offerings for HPC hardware, software and storage solutions at SC 18
  • Latest HPE-powered supercomputers continue to execute complex computational projects across academic, government and commercial use
  • HPE demonstrates first supercomputer open to astronauts on the International Space Station with the Spaceborne Computer

HPE bolsters HPC offerings with latest hardware and software developments

High performance computing (HPC) has become more essential to technological innovation, with broader industry impact than ever. According to a Hyperion Research study, 97 percent of companies that have adopted HPC said they could no longer compete or survive without it.[1] As HPC continues to evolve by integrating advanced processors, software, networking tools and more, it will become increasingly valuable to productivity and ROI, influencing widespread adoption across various market needs.

Industries such as manufacturing, financial services, energy, life sciences, academia and more are looking to HPC to address challenging workloads at a faster speed and unlock insights, accelerate product life cycle and enable advanced simulations and modeling. At HPE, we are fueling these opportunities with state-of-the-art, purpose-built HPC platforms to enable scalable and flexible high-performing compute power to support diverse applications and complex research.

  • HPE powers diverse HPC architectures in latest next-generation supercomputers with Arm and AMD

Taking Supercomputing to the Next Level

With HPC at its core, supercomputing has gained widespread use across academic, government, research and, in recent years, commercial organizations to execute complex computational projects that were inconceivable just a few years ago. These latest achievements in research include mapping the human genome, modelling the universe, creating complex deep neural networks and designing artificial intelligence (AI) models all of which HPE has supported through its leading HPC technologies.

We are committed to driving powerful supercomputers to make breakthrough discoveries and have just announced that HPE is powering the fastest supercomputer in the world for industrial production in collaboration with the High-Performance Computing Center of the University of Stuttgart (HLRS). The upcoming system, which HLRS has named Hawk, will be built on HPEs next-generation HPC platform, running new AMD EPYC processor code named Rome. Hawk will focus on computational engineering and research across science and industrial fields to advance applications in energy, climate, mobility and health.

Additionally, after unveiling Astra, built in collaboration with Sandia National Laboratories for the National Nuclear Security Administration (NNSA), the system has been recognized in this Novembers Top500 list as the worlds largest and fastest Arm supercomputer and the first Arm supercomputer to achieve 1.5 petaFLOPs. By introducing Arm to supercomputing, HPE is building a diverse network of HPC technologies to power next-generation systems. With Astra, HPE is also delivering memory-centric designs that support its customers data-intensive HPC workloads with world-class memory performance.

  • We bring robust, end-to-end HPC solutions to help achieve any business or organization goals

HPE at Supercomputing Conference 2018 (SC18)

In addition to our HPC platforms’ significant performance, we offer robust, end-to-end solutions designed to help any business or organization achieve its goals and deliver outcomes, whether that’s gaining a competitive advantage in the marketplace or solving some of the world’s toughest research problems. HPE has recently unveiled the following updates to its portfolio that will be showcased at SC18 at booth #2429 and in the Demo Zone at #2828:

      • Increasing Energy Efficiency in High-Performing Data Centers
        With the new HPE Adaptive Rack Cooling System, we are enabling customers to reach high performance levels to solve some of the most complex problems, without using additional energy. The system cools high-powered equipment to room temperature using a bi-directional, closed-loop cooling system to control air temperature at the heat source without any additional external cooling or containment. It packs up to 30 times the kilowatt capacity of a standard rack, enabling greater power density without reducing rack density within the data center.

      • Simplifying Data Storage and Management 
        HPE released the HPE Data Management Framework, a software solution for customers to optimize HPC data accessibility and storage resources through an automated, hierarchical, tiered storage management architecture.

      • Enhanced HPC Performance Cluster Manager 
        HPE enhanced the HPE Performance Cluster Manager, a fully-integrated system management solution for Linux®-based HPC clusters scaling up to 100,000 nodes. It enables accurate measurement and prediction of power usage for better capacity planning by collecting metrics. Latest improvements include better user experience with integration of the most popular ISV and open-source HPC workload management tools, system management, orchestration and monitoring tools. In addition, the HPE Performance Cluster Manager will support HPE HPC platforms, including the HPE Apollo 70 and HPE Apollo 35.

      • Advancing EDA Market Solutions with Cadence, Arm and Marvell
        Through a collaboration with Cadence, Arm and Marvell, HPE is empowering high-capacity data centers and enabling time-to-value for chip designers with a new EDA solution that integrates the Cadence Verification Suite with HPE Apollo 70, which is based on the Marvell® ThunderX2™ Arm processor. By deploying the HPE Apollo 70 systems with Cadence tools, customers gain a comprehensive verification environment with the flexibility to optimize compute capacity in diverse industries such as mobile, artificial intelligence, automotive, aerospace and defense, industrial and IoT sensor markets. This enhancement also maximizes on limited IT budgets by deploying up to 19 percent more cost-effective, higher throughput systems for improved farm utilization and better engineering productivity.

      • Transforming Data Centers with Intel AI and HPC Collaboration
        HPE and Intel have collaborated on the following developments, improving optimization of solution performance within data centers:
        • Select HPE ProLiant Gen10 Servers Gain FPGA Accelerators: The HPE ProLiant Gen10 DL360 and DL380 servers now support the Intel® Arria® 10 GX FPGA accelerator cards for enabling target workload acceleration such as database acceleration, financial back-testing of trading algorithms, and big data process acceleration.
        • Fast Interconnect Solution for HPC and AI: HPE enhanced its Apollo portfolio with the Intel® Omni-Path Architecture (Intel® OPA), enabling speed (100 Gb/s), low latency and cost-optimization to address entry-level to supercomputing clusters.

    • In-space Computing: First Above-the-Cloud Supercomputing Capabilities for Astronauts
      • We recently announced that since the Spaceborne Computer - the first commercial of-the-shelf supercomputer sent to the International Space Station (ISS) - successfully completed its one-year test experiment, proving it can turn on, achieve one teraFLOP, and withstand the harsh conditions of space.  It is now available for use as the first above-the-cloud supercomputer for astronauts. This is the first time researchers aboard the ISS can use the power of a supercomputer to process and complete research on their own, accelerating time-to-solution and testing boundaries for expanded space travel.
      • Demo Zone, HPE Booth #2828: HPE will demo the Spaceborne Computer and sharing live data feed on how it’s operating from space.
      • Birds of Feather Speaking Session: Additionally, Dr. Mark Fernandez, HPE payload software developer for the Spaceborne Computer project, will present key learnings from its one-year mission in the following Birds of Feather session at SC18:
        • Session: HPC in Space: An Update on Spaceborne Computer after 1+ Year on the ISS
        • Date: Wednesday, November 14, 2018
        • Time: 12:15 - 1:15 p.m. CST
        • Room: D174  


    To learn more about HPE’s high performance computing solutions, please visit our HPC offerings page.

    [1] Hyperion Research Study of HPC ROI, June 2018


HPE triples performance and enhances energy efficiency in new supercomputer for National Renewable Energy Laboratory (NREL)

Press Release

HPE delivers first above-the-cloud supercomputing services for astronauts to advance space exploration

Press Release

HPE Delivers World’s Largest Arm Supercomputer for U.S. Department of Energy (DOE)

Press Release