Lawrence Livermore official pushes shared compute and warns of overseas shift and security risks

Shankar Sundaram, deputy principal associate director for mission and engineering at Lawrence Livermore National Laboratory, told the Assembly the lab’s high‑performance computing helped redesign a clinical antibody in weeks during COVID‑19 and is now being used to run open protein‑folding models such as OpenFold 3. "We redesigned the antibody in 3 weeks," he said, using the example to illustrate how compute can shorten a discovery cycle that otherwise stretches months or years.

Sundaram urged state consideration of shared compute, automated labs and public‑sector resources so startups and university researchers are not locked out by compute cost or lab scale. He warned that innovation migration overseas could hollow California’s ecosystem—manufacturing, clinical trials and workforce follow investment—and said public compute and lab access would help preserve the state’s leadership. He also raised biosecurity concerns around rapidly engineered biological sequences and emphasized the need for defensive screening and coordinated mitigation.

Why it matters: Lawrence Livermore’s examples connect national‑lab capacity to state competitiveness and public‑health readiness. Sundaram recommended combining compute, lab automation and connected datasets and noted the state can send procurement signals (e.g., by funding wastewater surveillance or wearables) to shape markets.

Next steps: The committee and panelists flagged Cal Compute as a near‑term policy item for discussion; Sundaram and others suggested state funding for shared compute and laboratory measurement capacity to pair with federal standards and private investment.