Scientists at La Jolla Institute for Immunology (LJI) are eager to put a new computing tool to the test. The Institute recently acquired a NVIDIA GH200 Grace Hopper™ Superchip, which offers new ways to analyze and understand immune system and disease data.
“Laboratories at LJI are ready to flex hardware like this,” says Michael Scarpelli, Vice President of Information Technology and Core Services at LJI. “This will help supercharge the kinds of science we do here.”
As Scarpelli explains, the superchip is powerful because it combines AI and high-performance computing (HPC) for rapid analysis of very large datasets. For example, LJI scientists may use the superchip to examine genomic data and shed light on the root causes of disease. Scientists can also harness the superchip’s AI capabilities for “predictive modeling” of how new vaccines or therapies might help different patient groups.
LJI accessed the superchip thanks to a long-time partnership with Cambridge Computer, which delivers HPC/AI and Enterprise technologies from several manufacturers. The chip, manufactured by technology company NVIDIA, is named for late computer programming pioneer and mathematician Grace Hopper, Ph.D.
LJI researchers are well known for advancing immune system data curation and analysis. The Institute has established several valuable databases used by scientists worldwide: the Immune Epitope Database and Analysis Resource (IEDB), the Database of Immune Cell Epigenetics (DICE), the Cancer Epitope Database and Analysis Resource (CEDAR), and the Human Immunology Project Consortium (HIPC) Data Coordinating Center (ImmuneSpace).
Many LJI investigators use these databases and develop other data science tools to better understand infectious diseases, cancers, allergy, and more. For example, LJI Associate Professor Ferhat Ay, Ph.D., recently developed bioinformatics tools to reveal how the immune system’s B cells rely on a key protein to produce life-saving antibodies.
Scarpelli looks forward to seeing how LJI scientists take advantage of the NVIDIA GH200 Grace Hopper™ Superchip.
“When you’re doing data science, the hardware you’re on really does matter,” Scarpelli says. “LJI is fortunate to get early access to this kind of technology. Now it’s time to throw projects at it and see how far we can push it—let’s lead the charge.”
