- Purdue University’s supercomputing power has leveled up in strength and speed with the new Anvil supercomputer. These are the details.
Purdue University’s supercomputing power has leveled up in strength and speed with the new Anvil supercomputer. And the new capacity is going to allow Purdue to contribute to the nation’s 21st-century research agenda by simultaneously powering scientific computational and data-driven tools at multiple universities.
Having launched operations in November, Anvil is now at full capacity, funded by a $22 million grant from the National Science Foundation. And in its mission to serve as a resource for Purdue and outside institutions throughout the U.S., it already has facilitated work from over 20 institutions in its first four months of operation.
As evidenced by its use thus far, Anvil’s next-level strength and versatility promised to be a boon for Purdue researchers who have identified a wide range of potential uses for its computing power.
For example, Jonathan Poggie, professor of aeronautics and astronautics, is looking to tap into Anvil to speed up his study on the prediction and eventual control of the aerodynamic heating produced in flight at high Mach speeds. And Daniel Aliaga, associate professor of computer science, and his research group are exploring the use of Anvil’s GPU capability to help facilitate the creation of “what-if” design tools in digital city planning. These futuristic tools are going to enable urban planners worldwide to automatically integrate, process, analyze and visualize complex interdependencies among urban form, function, and the natural environment.
Plus new treatments for heart disease are the promise of a project that came to Anvil from out of state. Yinglong Miao, an assistant professor in the Department of Molecular Biosciences and Center for Computational Biology at the University of Kansas, works in advanced biomolecular modeling and drug discovery. And he is using Anvil’s power to accelerate his biomolecular simulations to unlock remedies for heart disease.
Anvil’s current strength is now in a category that challenges the imagination. And at peak performance, it can now operate at an astonishing 5.1 petaFLOPS (floating-point operations per second). A computer system with one petaFLOPS is capable of performing one quadrillion (1015) FLOPS. A one-petaFLOPS computer system can manage in just one second what would take an individual a calculation every second for 31,688,765 years. Multiply those individual calculations by a factor of 5.1, and the product is Anvil’s single-second peak capacity.
Anvil’s new power is going to significantly increase the computing capacity available to users of the NSF’s Extreme Science and Engineering Discovery Environment (XSEDE), in which Purdue has been a partner for the past 10 years. It will now be able to deliver over 1 billion computing core hours to XSEDE each year. And 10% of Anvil’s computing power will be made available for high-impact initiatives including research partnerships with industry.
Researchers can request access to Anvil via the XSEDE allocation process. And new allocations may be requested at any time by filling out a request form via the XSEDE user portal. Allocation on Anvil for class use and training may be requested via the education allocation process throughout the year. Plus large research allocations can be requested four times a year (see timetable) and are peer-reviewed quarterly.
Anvil consists of 1,000 nodes — which are individual computers that consist of one or more central processing units (CPUs) together with memory. And each node has two 64-core, third-generation AMD EPYC processors and will deliver over 1 billion CPU core hours to XSEDE every year.
Anvil’s nodes are interconnected with 100 Gbps Mellanox HDR InfiniBand. And the supercomputer ecosystem also includes 32 large memory nodes, each with 1 TB of RAM, and 16 nodes each with four NVIDIA A100 Tensor Core GPUs (graphics processing units) providing 1.5 petaFLOPS of single-precision performance to support machine learning and artificial intelligence applications. The research on Anvil will be able to leverage a diverse set of storage technologies, anchored by a 10-plus PB parallel filesystem, boosted with over 3 PB of flash disk. Novel workflows will benefit from block and object storage systems also supported by Anvil.
KEY QUOTES:
“Anvil is not only the largest capacity system Purdue has ever built, it’s the most diverse. Its various components are all integrated in one place with GPUs and large memory nodes complementing a cluster of 1,000 compute nodes. As Anvil grows, it is also able to take on much more heterogeneous workflows that are more common in the research happening today at Purdue.”
— Carol Song, senior research scientist for ITaP Research Computing and principal investigator and project director for Anvil
“Using CPUs this work could take months, if not years. With the GPUs on Anvil, we can run these simulations much faster. Instead of needing months, we just need a couple of weeks.”
— Yinglong Miao, an assistant professor in the Department of Molecular Biosciences and Center for Computational Biology at the University of Kansas
Featured image: Pictured with the Anvil supercomputer are from left to right: Rajesh Kalyanam, research scientist; Carol Song, senior research scientist; Xiao Zhu, senior research scientist; and Preston Smith, executive director, Purdue Research Computing.
(Credit: Purdue University photo/Vincent Walter)
