Recently, NVIDIA announced the acquisition of two Israeli AI companies, Run:AI and Deci.

Recently, NVIDIA announced the acquisition of two Israeli AI companies, Run:AI and Deci. According to the news, NVIDIA's acquisition of these two companies aims to reduce the cost of developing or running generative AI models, helping customers to use AI computing resources more effectively, thereby boosting the demand for NVIDIA's AI chips.

On April 24th local time, NVIDIA announced that it had signed a definitive acquisition agreement with Run:AI. The company believes that this will help customers use their AI computing resources more effectively.

NVIDIA did not disclose the specific terms of the transaction, but according to CTech, the deal is estimated to be worth about $700 million. It is reported that the acquisition of Run:AI is NVIDIA's largest acquisition in Israel since it acquired Mellanox for $6.9 billion in March 2019.

Public information shows that Run:AI was established in 2018, and it has developed a unique orchestration and virtualization software layer tailored to the specific needs of AI workloads running on GPUs and similar chips. Based on the Kubernetes AI cloud container platform, Run:AI can achieve efficient utilization of GPU cluster resources by automatically allocating the necessary computing power. In simple terms, its product can support developers to run multiple AI workloads in parallel, thereby improving the utilization efficiency of AI chips and reducing costs. Currently, the company has about 150 employees and has raised a total of $118 million in financing.

Advertisement

At present, Run:AI's solutions have been integrated with NVIDIA's DGX, DGX SuperPOD, Base Command, NGC containers, and AI Enterprise software products. Customers of NVIDIA DGX and DGX Cloud will be able to use Run:AI's AI workload capabilities.

NVIDIA stated that since 2020, Run:AI has been a close partner of NVIDIA. In the future, NVIDIA stated that it will continue to provide Run:AI's products with the same business model and will continue to invest in Run:AI's product roadmap as part of NVIDIA DGX Cloud.NVIDIA DGX Cloud, a product released by NVIDIA on March 21, 2023, is an AI supercomputing service that enables enterprises to instantly access the infrastructure and software required for training advanced models used in pioneering applications such as generative AI. NVIDIA DGX Cloud offers dedicated NVIDIA DGX AI supercomputing clusters, paired with NVIDIA AI software, allowing every business to access AI supercomputing through a simple web browser, eliminating the complexity of acquiring, deploying, and managing on-premises infrastructure. Enterprises can rent DGX Cloud clusters on a monthly basis, enabling them to quickly and easily scale the development of large, multi-node training workloads without waiting for the often high-demand accelerated computing resources.

Another AI company acquired, Deci, is also committed to "cost reduction and efficiency enhancement" of AI chips. According to The Information, NVIDIA has agreed to acquire the Israeli startup Deci, with the specific acquisition amount undisclosed.

Deci provides services to clients including Adobe and Applied Materials. It adjusts AI models to run more cost-effectively on AI chips. Initially, Deci assisted in enabling relatively simple AI applications on edge devices such as mobile phones and cars, then shifted to developing its own large models to compete in the open-source domain. Its self-developed models have been uploaded to its website and the Hugging Face community.

In December of the previous year, Deci's model DeciLM ranked first in the Hugging Face model rankings with 7 billion parameters, surpassing open-source competitors like Llama and Mistral. However, after Google launched the new Gemma model at the end of February, Deci lost its leading position.

Reports from The Information indicate that two individuals familiar with the transaction suggest that Run:ai's ability to improve AI chip efficiency may attract more customers to use NVIDIA DGX Cloud. Additionally, NVIDIA may offer Deci along with CUDA software, enabling developers to build AI-driven applications more cost-effectively.

NVIDIA also noted in its official press release that customers expect to benefit from better GPU utilization, improved GPU infrastructure management, and greater flexibility brought by an open architecture.

An analysis by data provider Dealogic shows that over the past year, NVIDIA has invested in more than 30 startups, more than three times that of 2022. Areas of interest for NVIDIA's investments include software, pharmaceuticals, and healthcare. In February of this year, NVIDIA stated in a filing with the U.S. Securities and Exchange Commission (SEC) that the company also holds shares in several small companies applying AI technology. These investments seem to have paid off. NVIDIA's financial reports indicate that as of January, the value of the company's investments is approximately $1.55 billion, compared to $300 million the previous year.

In this new wave of AI technology, chip giant NVIDIA is emerging as a powerful advocate for star AI companies. As the inventor of visual computing technology and Graphics Processing Units (GPUs), the rise of deep learning has made GPUs shine brightly, and NVIDIA is highly sought after. The investment list of the past one or two years can more directly reflect its ambitions in AI.