Douglas Jacobsen

Douglas Jacobsen designs and implements features and helps shape the roadmap for Slurm, drawing on more than a decade of running supercomputers from several perspectives. A bioinformatician by training (PhD, University of Michigan), he joined Lawrence Berkeley National Laboratory’s NERSC in 2012 to support science users, and rapidly moved into systems work on several generations of NERSC Cray supercomputers, including Hopper, Edison, Cori and Perlmutter. In 2015 he led NERSC’s migration to Slurm, and later worked to scale it to Cori’s roughly 13,000 nodes. In 2023 he moved to Azure to support cloud-based AI supercomputers for Microsoft’s most critical AI partners, then to Microsoft AI, where he was HPC technical lead for several AI training supercomputers. In 2026, he joined NVIDIA, where he continues to focus on making Slurm work for the researchers who depend on it across a variety of scales and use cases.