Led by chemistry professors Geoffrey Hutchison along with co-leads Lillian Chong in chemistry, David Koes in computational and systems biology, and Inanc Senocak in mechanical engineering and materials science, a multi-disciplinary team of Pitt researchers have landed a $1.2 million National Science Foundation award to purchase major new computational resources for Pitt's Center for Research Computing, and greatly boost the university's capabilities in speed, power, and scope. Matching internal funds from the university will create a total of over $1.5 million for new computational hardware.
The new resources will immediately benefit many groups in the chemistry department, and over 30 NSF-funded research groups across Pitt, representing over $18 million in research grants in fields including chemistry, computational biology, chemical engineering, materials science, psychology, astrophysics, weather forecasting, computer science, energy, and sustainability, among others.
The grant is not simply funding new computers. One cluster is based on new state-of-the-art Nvidia graphics processing units (GPU) that are two times faster than previous generation GPUs and 14-50 times faster than standard central processing units (CPU). GPU technology is driving a revolution in applying machine learning, data science, and artificial intelligence in every area of science and engineering. Each new node will provide 4 or 8 GPUs and at least 27,648 compute cores, vastly expanding the resources at Pitt for GPU computing.
“Not only are they advanced resources, but we’re going to have a lot of them,” says Hutchison. “Colleagues are surprised when I tell them how many GPUs we are getting.”
Another system will include state-of-the-art compute nodes for large memory CPU calculations. In many cases, researchers need to use many computers together, connected by a high-speed network, simply to hold large data sets in memory. The new cluster will include over 2,300 CPU cores and a total of 18 terabytes of system memory, enabling an immense range of scientific computing.
Beyond simple acceleration, the resources will enable transformative research with advanced molecular dynamics, new machine learning surrogates for quantum chemical calculations of molecular and materials energies and properties, rare-event sampling in protein folding and binding, fMRI neuroscience, vastly more accurate weather grids, and next generation digital astronomy.
The new expanded hardware will advance both undergraduate and graduate courses and educational experience across chemistry and a similarly broad range of departments and courses. These expanded resources will enable Pitt to expand a synergy between research and education at all levels, reaching beyond the university to faculty, staff, and students at Howard University, other historically black colleges and universities (HBCUs), and many undergraduate faculty and students both regionally and nationwide. The new resources will also expand scientific computing to students in the Pittsburgh Public School district, including nearby Pittsburgh Science and Technology Academy and Pittsburgh Public Allderdice, both urban schools with diverse student populations.