Nvidia announced on Monday that it has acquired SchedMD, a developer of open-source AI software, marking an effort to further strengthen its position in the burgeoning artificial intelligence sector. The move comes as the chip giant doubles down on open-source technologies to reinforce its ecosystem and defend against increasing competition.
Long renowned for its high-performance chips, Nvidia has steadily expanded its offering to include a suite of AI models that range from physics simulations to autonomous driving solutions. These models are used by both researchers and enterprises, in addition to the company’s widely adopted CUDA software platform—considered a key industry standard among developers. Software has become a critical asset for Nvidia, with CUDA playing an integral role in maintaining its lead in the AI market.
SchedMD, established in 2010 by Morris “Moe” Jette and Danny Auble in Livermore, California, provides the Slurm workload manager, an open-source solution for scheduling intensive computing jobs across data center servers. While the core technology is freely accessible to developers and companies, SchedMD generates revenue by offering engineering support and maintenance.
Financial details of the transaction were not disclosed. Nvidia has stated it plans to continue distributing SchedMD’s software as open source.
In a blog post, Nvidia highlighted that Slurm, which runs on the company’s latest hardware, is part of the critical infrastructure for generative AI—used to manage the training and inference of large foundational models, a need shared by model developers and AI practitioners.
Earlier the same day, Nvidia introduced a new line of open-source AI models, touting improvements in speed, cost, and intelligence compared to previous iterations. The launch takes place as the company faces increasing pressure from emerging open-source models developed by Chinese AI research labs.
SchedMD counts cloud infrastructure provider CoreWeave and the Barcelona Supercomputing Center among its clients.
