As Hyak, the UW’s supercomputer, is updated for faster speeds and GPU capability, more researchers are wooed by its value.
By Gretchen Konrady
Hyak means fast in the Chinook Wawa language and given the unrelenting pace of change in the world of computers, pushing the pedal to the metal is the only way to stay relevant in UW computing circles.
That is why UW-IT continues to enhance Hyak, the University’s supercomputer, to ensure it best serves ever-growing faculty and student research needs.
Upgrades include a new hardware infrastructure now underway, and recent expansions in GPU (Graphics Processing Unit) capacity. Hyak also has expanded its reach to lend its computing power to multiple research projects at UW Bothell and UW Tacoma.
Hyak supports a broad range of workloads — from traditional high-performance computing to data science approaches such as machine learning and other computation-heavy Big Data problems.
An infrastructure update underlies Hyak’s third-generation build, named Klone, which means “three.” The update, fully implemented in the fall of 2021, improved performance even while leveraging existing standard hardware, and included a new CPU cluster with more data storage overall, faster read/write capability and a more performant (twice as fast), cost-effective and scalable inter-networking architecture among Hyak’s processors.
Other cost-saving efficiencies in the update included migrating to standardized data center components — such as racks and electrical connections — in partnership with the UW-IT Data Center Services and Operations (DCSO) team.
“The hardware refresh will give us more flexibility and allow us to be hardware- and vendor-agnostic because we’re using open standard components,” said Nam Pho, director of cyberinfrastructure in UW-IT. “And that will make our operations and maintenance in the Data Center more agile.”
Although GPUs have been a part of Hyak since its inception, providing specialized ability to perform certain computations much faster than a traditional computer’s central processing unit, its GPU capacity has grown tenfold over the past three years to meet the growing demand.
The expansion has happened largely through funding and a partnership with the UW’s College of Engineering, which increased its capacity by 60 percent over the last year. This was driven by the growth in research of the Paul G. Allen School of Computer Science & Engineering (CSE) using the Klone GPU sub-system for deep learning and artificial intelligence projects.
“Though machine learning has historically been the biggest use case for GPUs on Hyak, researchers are beginning to see other GPU-accelerated applications, such as for molecular dynamics. Any process that relies on a lot of matrix multiplication is a candidate for GPU acceleration,” Nam said.
Hyak now being leveraged by Bothell and Tacoma researchers
Hyak’s power and potential — and its long list of successful research projects — has not gone unnoticed at UW Bothell and UW Tacoma, two smaller campuses where researchers have sought more access to major computing resources.
In September, UW Bothell received a $394,000 grant from the National Science Foundation that allowed the campus to become a member of the Hyak community — with dedicated research computing capacity for its faculty and students.
On their grant application, UW Bothell associate professor Eric Salathé, who conducts research on regional climate change and the impacts of climate change on human and natural systems, said, “It’s challenging for a small campus to sustain the continuity and economies of scale available to large research campuses where funded projects can aggregate resources in order to maintain innovative research programs and provide for new faculty.”
In particular, access to common campus research computing resources is an important aid in recruiting and retaining a diverse faculty engaged in multidisciplinary research — which has been proven time and time again with Hyak.
At UW Bothell, Hyak is expected to aid research across a wide range of computational research fields including astrophysics, climate modeling, biochemistry, genomics, machine learning, operations research, and mathematics. The new capacity also will allow for the development of a larger cross-campus initiative for computational science education supporting UW Bothell’s undergraduate teaching mission.
At UW Tacoma, Hyak is being brought onboard to provide computing and storage power for The Salish Sea Modeling Center (SSMC), a computational platform launched in June 2020 at the Center for Urban Waters.
The urban waters center is a joint project by UW Tacoma, the City of Tacoma and the Puget Sound Partnership — launched in 2010 to study and restore ecosystems across the Salish Sea, a vast area that includes Puget Sound, the Strait of Juan de Fuca and Georgia Strait in British Columbia.
How water moves and tides ebb and flow are key to understanding the Salish Sea, and the basis for SSMC’s work. With funding from the City of Tacoma, UW Tacoma purchased 520 cores and 50 TB of storage in Hyak to augment SSMC’s capabilities. In part, Hyak will host sophisticated hydrodynamic computer models that can help predict changes in the ecosystem.
One of them is the Salish Sea Model, an advanced computer simulator that “accurately describes how water, sediments, and nutrients enter and cycle through the Salish Sea.”
SSCM’s short-term goal is to make the Salish Sea Model and other computed solutions available over Hyak, making it easier for its own researchers to collaborate with others in interdisciplinary investigations, grants and projects as co-principal investigators.
“It’s great to see UW Tacoma and Bothell researchers put their trust in Hyak,” said Erik Lundberg, assistant vice president of UW-IT’s Research Computing and Strategy division. “Our job is to provide the tools that all UW researchers need to do their work, no matter what campus they’re working from. For some, Hyak will be a perfect solution, for others, they’ll gravitate to the cloud. We’re here to help them along the way.”
Learn more about Hyak and other computing resources
A new GPUs for Machine Learning page has been added to IT Connect to help guide researchers and educators toward the most appropriate place for doing their research and teaching. They can choose from Hyak or cloud, or if they have smaller jobs, they can use platforms such as Google Colab, or even special-purpose, GPU-based, machine-learning laptops.
Because students are relying more on GPUs for their research, a GPU-based supercomputing cluster prototype was developed last year by the student Research Computing Club and deployed with the help of UW-IT and funding from the Student Technology Fee. Access to Hyak is available to all students through the UW Research Computing Club.