The Hawaiʻi Data Science Institute at the University of Hawaiʻi was awarded $1 million by the National Science Foundation’s Major Research Instrumentation Program that will allow the university to acquire and deploy a new community computer cluster in support of computational and data-intensive research for the 10-campus system.

“This award brings together some of UH’s greatest strengths and important new initiatives, leveraging several investments and strategic decisions made over the past several years to help our faculty and students apply modern computational and visualization approaches to answer questions about the world around us,” said UH President David Lassner. “Housed in our sustainable new data center in the Information Technology Building, this state-of-the-art high-performance computing system will support some of our major new initiatives including our work to understand our water systems in Hawaiʻi, our study of the microbiome of our environment, and our new multi-disciplinary data science institute.”

The three initiatives Hawaiʻi Established Program to Stimulate Competitive Research’s ʻIke Wai project, the Center for Microbiome Analysis through Island Knowledge and Investigation and the Hawaiʻi Data Science Institute all conduct work with advanced computational needs that will benefit from the new computing cluster.

“The High-Performance Computing cluster will support and elevate the leading-edge work of our faculty and students across campus,” said UH Mānoa Provost Michael Bruno. “The fact that both faculty and students will have access to this free, shared resource is truly exciting and transformational.”

New computing cluster

The new cluster, named Mana, will add graphics processing unit (GPU) nodes for intensive computing among other upgrades to the UH high-performance computing (HPC) cluster and support 15 early career, pre-tenure faculty with research programs in data science.

UH Mānoa Information and Computer Sciences Assistant Professor Peter Sadowski explained how modern artificial intelligence (AI) and machine learning require GPU machines. “My research focuses on AI and machine learning for data science applications in precision health, oceanography, high-energy physics and astronomy,” Sadowski said. “The UH HPC is crucial to my work. It enables members of my group to routinely train deep neural networks on 50 GPUs simultaneously. This would be significantly more difficult and expensive on a commercial cloud service or a government supercomputer.”

Professor Jason Leigh, principal investigator of Mana, spoke of HPC’s importance for the future, “Mana will provide us with much needed computing power to realize the next-generation of artificial intelligence-enhanced data visualization, analysis and collaboration tools for accelerating research, development and workforce training, even in fields as diverse as cyber-security and creative media.”

“We are very excited to receive this award which recognizes the research excellence of all our faculty at UH and most importantly the enormous research potential of our new investigators,” said Gwen Jacobs, principle investigator and UH director of cyberinfrastructure. “Our long-term goal within the Hawaiʻi Data Science Institute is to provide leading-edge computational and data resources to insure that our new investigators will thrive at UH.”