6 of the top 10 fastest verified systems in the world and 10 of the top 20 most energy-efficient supercomputers in the world were built and delivered by HPE
November 28, 2025
IN THIS ARTICLE
· El Capitan is crowned the fastest supercomputer in the world again, marking the 8th consecutive time an HPE-built system has earned that distinction.
· Continuing to build upon Cray’s 50-year legacy of delivering energy efficiency through direct liquid cooling, HPE delivers 10 of the 20 most energy-efficient supercomputers in the world.
· Two HPE-built sovereign AI factories debut on the TOP500 list as the most powerful systems in Canada and Central Asia.
· The Chapel open source project joins the High Performance Software Foundation, expanding its community reach.
The world’s three fastest exascale supercomputers were built and delivered by HPE according to the recent TOP500 list, reflecting HPE’s global leadership in delivering high performance computing (HPC) systems built for the artificial intelligence (AI) era. HPE also continues to deliver industry-leading power efficiency with innovative direct liquid-cooling featured in 10 of the world’s top 20 most energy-efficient supercomputers, based on the Green500 list.
HPE is shaping the future of AI and HPC for years to come through a new portfolio and open source updates with far-reaching industry impact. Extending its supercomputing leadership, HPE announced the second generation of exascale supercomputers and a new portfolio of end-to-end solutions that include reimagined architecture, storage, compute and accelerator blades, networking, software, and direct liquid cooling. HPE is also committed to advancing open-source technologies that will accelerate industry-wide innovation in HPC and AI as the Chapel project joins a new open-source software foundation.
Performance leadership: HPE built 6 out of the 10 fastest verified supercomputers on the planet
For the eighth consecutive time, a system built and delivered by HPE achieved recognition as the fastest supercomputer in the world, as independently verified by the TOP500 list, which is updated twice yearly.
Notably, this is also the third time that HPE-built exascale supercomputers swept the top three spots. This achievement highlights HPE’s unmatched expertise in designing, building, and supporting large-scale infrastructure that accelerates scientific discovery, innovation, and enterprise transformation. The HPE-built systems among the top 10 of the TOP500 list are:
No. 1 – El Capitan
Built for Lawrence Livermore National Laboratory (LLNL), El Capitan retains its position as the world’s fastest supercomputer and achieved a groundbreaking 1.809 exaflops – a 4% performance gain over the system’s previous benchmarks. The system also continues to rank No. 1 on both the High-Performance Conjugate Gradient (HPCG) benchmark, which measures the performance of scientific applications, and the HPL-MxP benchmark, which tracks the convergence of HPC and AI with mixed-precision computations. For all its power, El Capitan is also one of the most energy-efficient systems in the world, ranking No. 23 on the Green500 list. Powered by AMD Instinct™ MI300A APUs, this HPE Cray Supercomputing EX system enables critical research in earth sciences, physics, materials science, and molecular biology.
No. 2 – Frontier
Constructed for Oak Ridge National Laboratory (ORNL) and powered by AMD Instinct MI250 GPUs and AMD EPYC CPUs, Frontier initially debuted as the first verified exascale system in 2022. Delivering 1.353 exaflops, Frontier continues to drive advancements in quantum physics, renewable energy, and precision engineering. Frontier also ranked No. 3 on the HPL-MxP benchmark tracking mixed-precision calculations for HPC and AI workloads, delivering 11.4 exaflops.
No. 3 – Aurora Built in collaboration with Intel for Argonne National Laboratory (Argonne), Aurora delivers 1.021 exaflops and is the second-most powerful system on the HPL-MxP benchmark, achieving 11.6 exaflops for mixed-precision calculations that are vital for AI and data-driven research.
No. 6: HPC6
Developed for Eni, one of the world’s largest energy tech companies, HPC6 is the most powerful on-premises supercomputer for industrial use. Eni leverages the system’s 477.9 petaflops to, among other things, optimize the company’s industrial processes, enhance fluid dynamics studies for CO₂ storage, and simulate plasma behavior for magnetic confinement fusion, supporting Eni’s decarbonization strategy through the advanced study of energy transition and materials science.
No. 8: Alps
Created for Swiss National Supercomputing Centre (CSCS), Alps delivers 434.9 petaflops to a broad community of researchers in Switzerland and around the world to study medicine, quantum chemistry, climatology, and AI/machine learning.
Made for EuroHPC JU, LUMI is a pan-European supercomputer with 379.7 petaflops of performance that supports climate modeling, neural network training in European languages, and diverse scientific disciplines across the continent.
Lasting legacy of energy efficiency: 10 of the top 20 energy-efficient supercomputers are built by HPE
HPE’s success in powering the majority of the world’s 10 fastest supercomputers is matched by the 50-year legacy of Cray in delivering energy efficiency through direct liquid cooling. Modern HPE systems feature 100% fanless direct liquid cooling systems architecture, which delivers industry-leading energy efficiency. Of the top 20 systems on the Green500 list, these 10 were built and delivered by HPE.
· No. 4 – Isambard-AI, developed for the University of Bristol, is a key part of the UK Government’s AI Research Resource (AIRR) that is intended to boost the country’s capabilities in responsible and cutting-edge AI development in fields such as robotics, big data, climate research, and drug discovery.
· No. 7 – SSC-24 Energy Module, created for Samsung Electronics, is the most energy-efficient supercomputer owned by an enterprise. The SSC-24 system also ranks No. 21 on the TOP500 list.
· No. 8 – Helios GPU, constructed for Academic Computer Centre Cyfronet AGH, is also Poland’s fastest supercomputer at No. 96 on the TOP500.
· No. 10 – Portage is a benchmarking system built by HPE to test real-world HPC and AI workloads and workflows for the company and its customers. Portage is also No. 70 on the TOP500 list of the world’s fastest supercomputers.
· No. 12 – Hunter is built for the High-Performance Computing Center of the University of Stuttgart (HLRS), one of Germany’s three national high-performance computing centers. Hunter supports a wide range of applications in engineering and applied sciences for industry and academia.
· No. 16 – RZAdams, built for LLNL, is an additional sibling system for the world’s fastest supercomputer, El Capitan. The RZAdams system also ranks No. 72 on the TOP500 list.
· No. 17 – Frontier TDS is a test and development system built for ORNL which also ranks among the 100 fastest supercomputers in the world as per the TOP500.
· No. 18 – Shaheen III – GPU is a system developed for King Abdullah University of Science and Technology (KAUST) to support the Kingdom of Saudi Arabia’s scientific excellence, research, development, and innovation priorities. At No. 18 on the TOP500 list, Shaheen III – GPU is also the most powerful supercomputer in the Middle East.
· No. 19 – Tuolumne was built for LLNL as a companion system to El Capitan and is dedicated to open science. Tuolumne also ranks No. 12 on the TOP500 list of the world’s fastest supercomputers.
· No. 20 – El Dorado, built for Sandia National Laboratories, is an application-readiness test system that supports the United States’ stockpile stewardship mission.
New & notable debuts
Over the past year, HPE has built and delivered multiple new AI and HPC systems enabled by services that provide customers with turnkey implementation. As a result, some new and notable sovereign AI systems have debuted on the TOP500 and Green500 in the past 6 months. · TELUS sovereign AI factory takes Canada’s top spot: Newly launched in September, this system provides Canadian-based businesses, researchers, and public sector organizations access to advanced AI solutions that are fully developed, managed, and hosted within the country to ensure privacy and data sovereignty. Its 22.74 petaflops of performance officially ranks it as the fastest, most powerful system in Canada.
· Alem.Cloud Sovereign AI debuts as Central Asia’s fastest regional system: Located in the capital city of Astana, Alem.Cloud is operated by the Ministry of Artificial Intelligence of the Republic of Kazakhstan and plays a central role in advancing the region’s HPC capabilities. The system entered the TOP500 list at No. 86, marking a historic milestone for Kazakhstan. Its 20.48 petaflops of performance make it the most powerful system in the country. Alem.Cloud provides secure and trusted digital infrastructure that supports government services, universities, startups, local IT companies and underpins the development of domestic AI models, including large language models in the Kazakh language. Additionally, HPE’s recent system wins include Discovery and Lux, Mission and Vision, Janus and a new system focused on quantum error correction that is expected to advance hybrid quantum-classical computing.
Impactful science
Supercomputers and industry-leading AI clusters are powerful systems designed to address some of the world’s most complex challenges.
HPE plays a critical role in helping organizations and researchers make scientific breakthroughs, solve engineering problems, and change the way people live. Some recent notable projects include the following ACM Gordon Bell Prize finalists:
· Researchers used El Capitan to generate a massive library of physics-based earthquake and tsunami simulations, creating a high-fidelity digital twin of the seafloor. This study enables scientists to more accurately predict tsunami behavior and could lead to the development of a faster, more reliable early warning system for coastal communities.
