AI giant companiesNvidiaContinuing its strategic acquisition drive, it announced today that it will acquire an Israeli startup to make AI chips more efficient. The chipmaker has signed a "definitive agreement" to acquire Kubernetes-based software provider Run:aiThe latter helps optimize AI applications and workloads on the graphics processing unit (GPU).
The deal will be close to $700 million, although the exact amount has not been disclosed, people familiar with the matter told TechCrunch. Previous discussions pegged the purchase price tag at a much higher $1 billion. The deal marks a series of strategic moves and investments by NVIDIA'sup to dateinitiatives, giving it an increasingly dominant position in the AI technology stack.
Run:ai helps organizations manage and optimize their compute infrastructure, whether in the cloud, locally or in hybrid environments. Its orchestration and virtualization software layer specifically targets AI workloads running on GPUs and other chipsets.
The company's centralized interface allows users to manage shared compute infrastructure. Developers can pool GPUs for a variety of tasks and share compute resources - this can be "portions of a GPU" or multiple GPUs, or GPU nodes running on different clusters. Customers benefit from better GPU cluster resource utilization, improved infrastructure management, and greater flexibility.
Run:ai can "slice and dice" and "dynamically allocate" GPUs, as well as combine and manage all workflows and data streams. The company has built its open platform on Kubernetes, supporting all Kubernetes variants and integrating with third-party AI tools and frameworks.
Run:ai's capabilities will be extended to NVIDIA DGX and DGX Cloud customers, and NVIDIA will continue to offer its products under the "same business model" "in the near future."
This is not a NVIDIAFirstThe acquisition comes after the company has made more than a dozen previous acquisitions. Notably, NVIDIA paid $6.9 billion to acquire high-performance computing company Mellanox in 2019, and has also acquired companies such as OmniML for edge AI workloads, SwiftStack for data storage and management, and Excelero for block storage. The company also made several other investments in hardware, software, data center management platforms, robotics, security analytics and mobile capabilities.
The acquisition demonstrates NVIDIA's control of the AI technology stack, showcasing aggressive expansion of the entire AI ecosystem and securing future revenue streams. As NVIDIA builds strong partnerships with all major AI and cloud service providers, the company continues to announce new innovations, including the recent launch of its "gigantic GPU" and multimodal AI project GR00T.
In response, one Twitter user said that integrating Run:ai into NVIDIA's existing DGX Cloud helps demonstrate that NVIDIA is vertically integrating its platform from chip to inference, essentially making it a one-stop shop for your AI needs. Another user shared a chart of NVIDIA's investments over the past four years and commented that the company is capitalizing on the current momentum to expand its ecosystem and secure future revenue streams. Startups (customers) depend on their GPUs and NVIDIA's growth depends on these startups.
Run:ai was founded in 2018 by Omri Geller and Ronen Dar. The company launched in 2019 with a $13 million investment and has since raised more than $105 million. run:ai has been working with NVIDIA for several years, with its products integrated into DGX, DGX SuperPOD, Base Command, NGC containers, and AI enterprise software, and its customers include Sony, Adobe, and BNY Mellon Bank of New York Mellon, among others.
"Run:ai has been working closely with NVIDIA since 2020, and we're both passionate about helping our customers get the most out of their infrastructures," Geller said in a NVIDIA blog post announcing the deal.