NVIDIA has completed the acquisition of Run:ai, an AI infrastructure startup. This deal means that Run:ai's software, which is currently only compatible with NVIDIA products, will be open-sourced, allowing competitors like AMD and Intel to adapt the software for their own hardware.

Following the merger, Run:ai stated that it plans to expand its team and increase market penetration. The open-source software will make the tool accessible to the entire AI ecosystem. The acquisition, announced in April, was reported to be valued at $700 million. However, due to concerns that the acquisition might negatively impact competition, the U.S. Department Justice and the European Commission launched investigations into the deal.

In December, the European Commission ultimately approved the transaction. Run:ai's platform can optimize performance, reducing the number of graphics processing units (GPUs) needed to run AI workloads and enabling AI infrastructure to handle up to ten times more work than before. By automatically allocating workloads to maximize hardware efficiency and prevent memory conflicts and resource contention, Run:ai enhances the efficiency of GPU clusters.

Additionally, Run:ai offers a suite of developer tools, including large language models, frameworks, and other components, to help software teams build their own neural networks and quickly deploy them on GPUs to create development environments.

Shortly after the acquisition announcement, antitrust regulators in the U.S. and EU began reviewing the deal, concerned that NVIDIA might use the acquisition to stifle technologies that could challenge its primary profit sources. To address these concerns, NVIDIA plans to open-source Run:ai's platform after the acquisition, which should alleviate worries about software access restrictions and suggest future support for third-party AI chips.

NVIDIA also intends to further enhance the functionality of the Run:ai platform, expand its market reach, and hire additional staff to support this process.