Argonne National Laboratory, a multidisciplinary science and engineering research centre operated by the University of Chicago for the U.S. Department of Energy, has installed Graphcore’s Intelligence Processing Units (IPUs) within its AI Testbed.
The AI Testbed, operated by the Argonne Leadership Computing Facility (ALCF), enables researchers to explore next-generation machine learning applications and workloads to advance the use of AI for science. Technologies selected for the testbed, such as Graphcore’s IPUs, complement the facility’s current and next-generation supercomputers to provide a state-of-the-art environment that supports pioneering research at the intersection of AI, big data, and high-performance computing (HPC).
The systems in the ALCF AI Testbed are purpose-built for machine learning and data-centric workloads, making them well suited to address challenges involving the increasingly large amounts of data produced by supercomputers, light sources, and particle accelerators among other powerful research tools. In addition, the testbed allows researchers to explore novel workflows that combine AI methods with simulation and experimental science to accelerate the pace of discovery. Testbed resources are available to members of the broader research community with proven problems to solve.
“The goal of the ALCF AI Testbed is to understand the role AI accelerators can play in advancing data-driven discoveries, and how these systems can be combined with supercomputers to scale to extremely large and complex science problems,” said ALCF Director Michael Papka. “As one of a handful of the promising technologies that we believe will play a large role in the future of scientific computing, Graphcore’s IPU was a good choice for inclusion in the AI Testbed.”
The IPU is a completely new kind of massively parallel processor for fine-grained, high-performance computing, specifically designed alongside Graphcore’s Poplar SDK to accelerate AI, machine intelligence, and HPC workloads. The IPU’s unique architecture offers proven performance and efficiency advantages over today’s CPUs and GPUs, empowering researchers to undertake entirely new types of work and solve complex, important problems that would otherwise be considered intractable.
An IPU system—installed by Dell, a Graphcore Elite Partner—has been in use at the ALCF since they began building the AI Testbed in 2021. The system is currently in use by Argonne for specific needs-based projects, and it will be opened to the larger user community at a later date.
“Our evaluation of the IPU’s performance on science workloads so far has yielded impressive results,” said Venkat Vishwanath, ALCF's Data Science Team Lead. “We are particularly impressed by the maturity of the Poplar SDK, and we are excited by the prospect of enhancing the AI Testbed even further with the addition of Graphcore’s latest hardware.”
Graphcore and Argonne are also working together on a range of research and benchmarking projects, more details of which will be announced in the coming months.