Nvidia has bought the Israeli startup Run:ai, which helps developers and operations teams manage and optimize infrastructure for artificial intelligence.
The financial terms of the transaction were not disclosed. According to TechCrunch sources, the acquisition amount was $700 million.
Earlier, the media reported on the possible valuation of Run:ai at $ 1 billion in case of successful negotiations with Nvidia. The dialogue between the parties probably went smoothly, although the final price could change.
Nvidia stated that it will continue to offer Run:ai products based on the existing business model and invest in their development within the Nvidia DGX Cloud AI framework. This platform provides corporate clients with computing infrastructure and software for training AI models.
Users of Nvidia DGX servers, stations, and the DGX Cloud will have access to Run:ai capabilities to deploy artificial intelligence workloads, especially generative ones, in multiple data centers.
"Run:ai has been working closely with Nvidia since 2020, and we are united in our desire to maximize the efficiency of infrastructure use by our customers," said Omri Geller, CEO of Run:ai.
Geller and Ronen Dar founded Run:ai in 2018 after studying at Tel Aviv University with Professor Meir Feder, the startup's third co-founder. They sought to create a platform for parallel "distribution" of AI models to various hardware locally, in the clouds or on the periphery.
Run:ai has few direct competitors, but other companies are also applying the concept of dynamic resource allocation to AI tasks. For example, Grid.ai develops software for parallel training of models on graphics accelerators, processors and other equipment.
In a short period of time, Run:ai has managed to gain a wide client base from Fortune 500 companies and attract venture capital investments. Before the acquisition, the startup raised $118 million from funds like Insight Partners, Tiger Global, S Capital and TLV Partners.
According to Nvidia Vice President Alexis Bjorlin, demand for the efficient use of AI computing is growing in response to the increasing complexity of deploying models. The ClearML survey showed that the key obstacles to scaling artificial intelligence in 2024 are the limited and high cost of computing resources, as well as infrastructure problems.
Run:ai is one of Nvidia's largest acquisitions since it bought Mellanox for $6.9 billion in March 2019.
Recall that in October 2023, Nvidia and Foxconn entered into a partnership for the "industrial AI revolution."
Subscribe to ForkLog news on Telegram: ForkLog AI — all the news from the world of AI!