HPC6

The HPC6 supercomputer, developed by Eni S.p.A. and built by Hewlett Packard Enterprise (HPE), is a highly advanced supercomputing system tailored for energy exploration and research, including seismic processing and reservoir modeling. Ranked 5th in the November 2024 TOP500 supercomputer rankings, it is currently the most powerful enterprise-focused supercomputer globally and the fastest in Europe, primarily focused on advancing energy exploration and sustainability research. It is an evolution of Eniโ€™s previous supercomputers, including HPC4 and HPC5, which were among the top-ranking systems in the TOP500 list of global supercomputers. Below are its key technical and architectural details:

Technical Specifications

  1. Performance:
    • HPL Benchmark (High-Performance LINPACK): Achieves 477.9 petaflops, ranking it 5th globally in the November 2024 TOP500 list.
    • Core Count: Includes a vast array of processors optimized for heavy computational workloads. It uses both CPU and GPU resources effectively for high throughput: 213,120 (3,330 ร— 64-core Optimized 3rd Generation EPYC 64C @2.0 GHz) CPU Cores and 2,930,400 (13,320 ร— 220 AMD Instinct MI250X) GPU Cores for a total of 3,143,520 Cores.
  2. Architecture:
    • Built using the HPE Cray EX235a architecture, similar to the Frontier supercomputer at Oak Ridge National Laboratory, which is the second most powerful globally. This architecture emphasizes high parallelism, scalability, and energy efficiency.
    • Incorporates a high-density integration of CPUs and GPUs optimized for large-scale, parallel workloads.
  3. Processors:
    • Employs advanced processors and accelerators, AMD EPYC CPUs and AMD Instinct GPUs, known for their high-performance capabilities in HPC environments.
    • These processors provide high computational throughput and energy efficiency.
  4. Networking:
    • Utilizes the Cray Slingshot-11 interconnect, enabling high-speed, low-latency communication crucial for parallel workloads.
  5. Software:
    • Uses Linux (RHEL 8.9) as its operating system.
    • Incorporates AI and machine learning frameworks to support simulation and optimization tasks.
  6. Energy Efficiency:
    • Installed at Eniโ€™s Green Data Center in Ferrera Erbognone, Italy, known for its sustainability measures and efficient power usage, emphasizing Eniโ€™s commitment to reducing carbon footprints.
    • Employs liquid cooling technology for enhanced energy efficiency with an energy consumption of only 8,461 kW.

Primary Applications

  1. Energy Transition and Sustainability:
    HPC6 is central to Eni’s research into renewable energy, energy storage, and carbon capture technologies.
  2. Oil and Gas Exploration:
    Enables advanced simulations for reservoir modeling, seismic data analysis, and fluid dynamics, improving exploration accuracy.
  3. AI and Computational Modeling:
    Supports large-scale AI-driven analytics and predictive modeling, aiding in process optimization and resource management.
  4. Scientific Research:
    Used for climate and environmental studies, including high-resolution modeling of global weather patterns and energy systems.

HPC6 represents Eniโ€™s strategic investment in supercomputing to drive forward the global energy transition, combining HPC and AI to tackle complex challenges in energy sustainability, climate modeling, and resource optimization.

See more details on: