Nvidia is buying Run:ai, an AI infrastructure management firm based in Tel Aviv. The deal’s specifics are not public, but it’s rumored to be around $700 million.
Earlier, CTech reported that Nvidia might pay over $1 billion for Run:ai. The negotiation process was smooth, except for a potential price adjustment.
Nvidia plans to keep selling Run:ai’s products and invest in its product roadmap. This is part of Nvidia’s DGX Cloud AI platform, which provides businesses with the infrastructure and software needed to train AI models. Nvidia’s DGX server, workstation, and cloud customers will also benefit from Run:ai’s capabilities, especially for generative AI deployments across multiple data centers.
Run:ai’s CEO, Omri Geller, expressed excitement about joining Nvidia. Geller, along with Ronen Dar and Professor Meir Feder, co-founded Run:ai. Their goal was to create a platform that could split AI models into fragments that run in parallel across different hardware.
While Run:ai has few direct competitors, other companies like Grid.ai are also focusing on dynamic hardware allocation for AI workloads. Run:ai has managed to attract a large customer base of Fortune 500 companies early on, which led to VC investments. Before the acquisition, Run:ai had raised $118 million from investors like Insight Partners, Tiger Global, S Capital, and TLV Partners.
Alexis Bjorlin, Nvidia’s VP of DGX Cloud, noted in a blog post that AI deployments are becoming more complex. A survey by ClearML found that the biggest challenge in scaling AI in 2024 has been computing limitations, followed by infrastructure issues.
Bjorlin said that managing and orchestrating AI workloads requires sophisticated scheduling to optimize performance. Nvidia’s accelerated computing platform and Run:ai’s platform will continue to support a broad ecosystem of third-party solutions. Together with Run:ai, Nvidia will enable customers to access GPU solutions anywhere.
Run:ai is one of Nvidia’s largest acquisitions since it bought Mellanox for $6.9 billion in March 2019.
