Nvidia Purchases AI Workload Management Startup Run:ai for $700 Million, According to Sources

Nvidia is set to acquire Run:ai, a Tel Aviv-based company specializing in optimizing and managing AI hardware infrastructure for developers and operations teams. While the financial details of the deal remain confidential, insiders reveal that the acquisition cost is around $700 million. Reports from CTech earlier today hinted at "advanced negotiations," suggesting Nvidia might have considered a price exceeding $1 billion, indicating successful discussions with only minor potential adjustments.

Nvidia plans to maintain Run:ai’s products “under the same business model” and will enhance its product roadmap as part of Nvidia’s DGX Cloud AI platform. This platform provides enterprise clients with access to vital compute infrastructure and software for training models used in generative AI and other applications. Customers utilizing Nvidia's DGX servers, workstations, and DGX Cloud will now benefit from Run:ai’s capabilities, especially for managing generative AI deployments across multiple data centers.

"Run:ai has been a close collaborator with Nvidia since 2020, sharing a commitment to empowering our customers to optimize their infrastructure," said Omri Geller, CEO of Run:ai. "We’re excited to join Nvidia and to continue this journey together."

Geller co-founded Run:ai alongside Ronen Dar after they studied under Professor Meir Feder at Tel Aviv University. Their vision was to create a platform that could decompose AI models into smaller fragments that could run in parallel, either on-premises, via public clouds, or at the edge.

Although Run:ai faces minimal direct competition, several firms are pursuing similar dynamic hardware allocation strategies for AI workloads. For instance, Grid.ai provides software enabling data scientists to train AI models across various processors and GPUs simultaneously.

Despite being relatively new, Run:ai has secured a significant customer base among Fortune 500 companies, attracting considerable VC investments along the way. Before this acquisition, Run:ai successfully raised $118 million from notable investors, including Insight Partners, Tiger Global, S Capital, and TLV Partners.

Furthermore, in a blog post, Alexis Bjorlin, Nvidia’s VP of DGX Cloud, emphasized the increasing complexity of customer AI deployments and the growing need for companies to enhance their utilization of AI computing resources. A recent survey conducted by ClearML, a machine learning model management firm, identified compute limitations—regarding availability and cost—as the primary challenge in scaling AI in 2024, followed closely by infrastructure issues.

“Managing and orchestrating generative AI, recommendation systems, search engines, and other workloads demands advanced scheduling to optimize system-level performance and utilize infrastructure effectively,” Bjorlin explained. “Nvidia’s accelerated computing platform combined with Run:ai’s solutions will continue to support a diverse ecosystem of third-party options, granting customers freedom and flexibility. Together with Run:ai, Nvidia will empower customers with a unified framework to access GPU solutions wherever needed.”

This acquisition marks one of Nvidia’s largest since acquiring Mellanox for $6.9 billion in March 2019.