Case Study: Argonne National Laboratory achieves award-winning AI-driven virus-variant tracking with Altair

Argonne National Laboratory Researchers Win Gordon Bell Special Prize

Argonne National Laboratory, a U.S. Department of Energy multidisciplinary research center operating the Polaris supercomputer, faced the urgent challenge of tracking rapidly evolving SARS‑CoV‑2 variants more quickly than traditional, labor‑intensive methods allowed. To scale and orchestrate GPU‑heavy AI workloads in advance of exascale, Argonne collaborated with Altair and used Altair PBS Professional for efficient workload management.

Using Altair PBS Professional to orchestrate GPU resources on Polaris alongside Cerebras and NVIDIA accelerators, Argonne trained a genome‑scale language model (GenSLM) and analyzed 1.5 million high‑quality SARS‑CoV‑2 genomes. The work won the 2022 Gordon Bell Special Prize, delivered much faster detection and characterization of variants of concern, and accelerated scientists’ ability to inform vaccine and treatment responses—helping to save lives worldwide.

Argonne National Laboratory

Bill Allcock

Director of Operations