Nvidia has finalized its acquisition of Run:ai, bringing an Israeli AI infrastructure management company deeper into its business at a time when GPU systems and cloud-based AI computing are central to how organizations build and run AI.
The deal was first announced in April. Run:ai, founded in 2018 by Omri Geller and Ronen Dar, has worked closely with Nvidia since 2020 and focuses on software that helps organizations manage AI computing resources more efficiently.
What Nvidia is getting with Run:ai
Run:ai’s core business is AI resource management. Its main product, the Atlas platform, helps organizations distribute computing power across multiple AI tasks running at the same time.
In practical terms, that means Run:ai sits in the layer between AI workloads and the infrastructure needed to run them. For companies using large GPU systems, the challenge is not only having enough computing power. It is also making sure that power is allocated where it is needed, when it is needed.
Run:ai has served a customer base that includes Fortune 500 companies and startups. The source article identifies finance, automotive, and healthcare among the sectors using the company’s technology.
Under Nvidia’s ownership, Run:ai said it will continue helping customers optimize AI infrastructure and GPU systems across several deployment models:
- on-premises systems
- cloud environments
- Nvidia’s DGX cloud service
That makes the acquisition relevant not just to hardware buyers, but also to companies trying to coordinate AI workloads across different infrastructure setups.
The open-source shift is the biggest change
The most notable post-acquisition change is Run:ai’s plan to make its software open source. Before this change, the platform worked only with Nvidia GPUs.
According to the source article, open-sourcing the software could potentially allow the platform to support other hardware platforms in the future. That point matters because AI infrastructure is often built from tightly connected software and hardware decisions.
If the software becomes more open, organizations may gain more flexibility in how they manage AI workloads. The source does not say when broader hardware support might arrive, or which platforms could be supported. The important confirmed fact is narrower: Run:ai is moving from a previously proprietary software stack toward open source.
For Nvidia, this creates an interesting balance. The company is acquiring a software platform that has been tied to its own GPUs, while also allowing that platform to move in a direction that could become less hardware-specific over time.
Why AI infrastructure management matters
AI systems can require many tasks to run in parallel. Training, tuning, testing, and deployment can all compete for computing resources inside the same organization.
Run:ai’s Atlas platform is described as a tool that takes an organization’s computing power and automatically distributes it across different AI tasks. That makes it a management layer for AI computing, not just another application running on top of infrastructure.
This kind of software becomes more important as organizations operate AI systems across multiple environments. Some may keep systems on-premises. Others may use cloud infrastructure. Some may use Nvidia’s DGX cloud service. Run:ai’s stated role under Nvidia is to keep helping customers optimize across those setups.
The acquisition also shows how AI infrastructure has become a software problem as much as a hardware problem. GPUs provide the computing power, but organizations still need tools to allocate that power efficiently and avoid bottlenecks between competing workloads.
The Israel connection and deal context
Nvidia has not officially disclosed the purchase price for Run:ai. Israeli business publication Calcalist reported in April that the deal was worth around $700 million, via Bloomberg.
Run:ai had previously raised a total of $118 million in funding. Its most recent Series C round was $75 million and was led by Tiger Global and Insight Partners.
The acquisition is Nvidia’s second major investment in Israel, following its $7 billion purchase of Mellanox Technologies in 2020. The source article also notes that the Mellanox deal is facing fresh scrutiny. Chinese state television reports that authorities are investigating Nvidia for potential antitrust violations related to that acquisition, even though China initially approved it in 2020.
The Run:ai purchase therefore lands in a broader context. Nvidia is expanding its AI infrastructure portfolio and its presence in Israel, while an earlier major acquisition in the same country is now linked to reported regulatory attention in China.
What to watch next
The clearest immediate takeaway is that Run:ai will continue operating around AI infrastructure optimization under Nvidia. Its customers, including Fortune 500 companies and startups, are expected to remain focused on getting more efficient use from AI computing resources.
The larger question is how the open-source plan changes the platform over time. The source article says the move could potentially allow support for other hardware platforms in the future, but it does not confirm a timeline or technical roadmap.
For now, Nvidia has added a specialized AI infrastructure management company to its portfolio. Run:ai brings software for distributing AI workloads across GPU systems, and its shift toward open source may become the most consequential part of the deal.