Hewlett Packard Enterprise delivers the exascale Aurora supercomputer to Argonne National Laboratory, achieving a new milestone in computing power and AI capabilities.
Hewlett Packard Enterprise (HPE) has taken another monumental step in the field of supercomputing with the delivery of the Aurora supercomputer to the United States Department of Energy’s Argonne National Laboratory. Named Aurora, this supercomputer has achieved an impressive milestone by reaching 1.012 exaflops, securing its place as the second-fastest supercomputer in the world according to the TOP500 list.
What sets Aurora apart is not just its speed but its classification as an exascale system. Exascale computing refers to systems capable of performing at least one quintillion (1,000,000,000,000,000,000) operations per second. This immense computational power is pivotal in solving some of the most complex and intricate challenges in science and technology.
Aurora, built in cooperation with Intel, represents a significant innovation in HPE’s portfolio as it is the second exascale supercomputer developed by the company, and notably, the largest AI-capable deployment globally. This milestone was celebrated during ISC High Performance 2024, a prominent annual event that focuses on high performance computing, networking, and storage.
In terms of technical specifications, Aurora is developed around the HPE Cray EX supercomputer architecture which is engineered precisely to support the demands of exascale computing. The system features a sophisticated network based on HPE’s Slingshot, an Ethernet-based supercomputing interconnect, which is crucial for enabling high-speed networking across its various components. The supercomputer comprises over 75,000 compute node endpoints, 21,248 Intel Xeon CPU Max Series processors, and 63,744 Intel Data Center GPU Max units – making it one of the largest GPU clusters around the globe.
In essence, Aurora is tailored to enhance Artificial Intelligence (AI) capabilities from the ground up. It empowers researchers to utilize advanced generative AI models which can significantly fast-track scientific discovery across multiple disciplines. Argonne’s early research using Aurora has already explored areas such as brain mapping, high-energy particle physics, and accelerated drug discovery – all enhanced by deep learning and AI technologies.
The achievement of Aurora is not just a technical feat but also a shining example of effective collaboration between the public and private sectors. The joint effort between HPE, Intel, the U.S. Department of Energy, and Argonne National Laboratory underlines the importance of combined expertise and resources in pushing the boundaries of what modern computing can achieve. Such partnerships are essential in driving forward scientific progress and innovation.
As the world increasingly leans towards intricate computational needs, especially in domains requiring vast AI capabilities, supercomputers like Aurora are crucial. They are not just machines of computational power but are pivotal tools in modeling complex phenomena, accelerating scientific research, and providing solutions to global challenges that were once thought insurmountable.
As stated by Trish Damkroger, senior vice president and general manager at HPE, the development of Aurora is a “significant milestone in exascale” that furthers the ability to make groundbreaking scientific discoveries and tackle some of the toughest problems faced around the world.
This significant advancement in HPE’s supercomputing technology underscores the company’s ongoing commitment to developing high-performance computing solutions that pave the way for new scientific achievements, further enhancing our understanding of both the digital and physical worlds.