NVIDIA in the present day introduced it has acquired SchedMD — the main developer of Slurm, an open-source workload administration system for high-performance computing (HPC) and AI — to assist strengthen the open-source software program ecosystem and drive AI innovation for researchers, builders and enterprises.
NVIDIA will proceed to develop and distribute Slurm as open-source, vendor-neutral software program, making it extensively accessible to and supported by the broader HPC and AI group throughout numerous {hardware} and software program environments.
HPC and AI workloads contain advanced computations working parallel duties on clusters that require queuing, scheduling and allocating computational assets. As HPC and AI clusters get bigger and extra highly effective, environment friendly useful resource utilization is vital.
Because the main workload supervisor and job scheduler in scalability, throughput and sophisticated coverage administration, Slurm is utilized in greater than half of the highest 10 and prime 100 techniques within the TOP500 record of supercomputers.
Slurm, which is supported on the newest NVIDIA {hardware}, can be a part of the vital infrastructure wanted for generative AI, utilized by basis mannequin builders and AI builders to handle mannequin coaching and inference wants.
“We’re thrilled to hitch forces with NVIDIA, as this acquisition is the final word validation of Slurm’s vital position on the earth’s most demanding HPC and AI environments,” stated Danny Auble, CEO of SchedMD. “NVIDIA’s deep experience and funding in accelerated computing will improve the event of Slurm — which is able to proceed to be open supply — to fulfill the calls for of the subsequent era of AI and supercomputing.”
NVIDIA has been collaborating with SchedMD for over a decade and can proceed investing in Slurm’s improvement to make sure it stays the main open-source scheduler for HPC and AI.
NVIDIA will speed up SchedMD’s entry to new techniques — permitting customers of NVIDIA’s accelerated computing platform to optimize workloads throughout their whole compute infrastructure — whereas additionally supporting a various {hardware} and software program ecosystem, so prospects can run heterogeneous clusters with the newest Slurm improvements.
NVIDIA will proceed to supply open-source software program assist, coaching and improvement for Slurm to SchedMD’s a whole bunch of consumers, which embrace cloud suppliers, producers, AI firms and analysis labs spanning industries equivalent to autonomous driving, healthcare and life sciences, vitality, monetary companies, manufacturing and authorities.
Along with SchedMD, NVIDIA is bolstering the open-source software program ecosystem to catalyze HPC and AI innovation throughout industries, at each scale.
