skip to main content
Close Icon We use cookies to improve your website experience.  To learn about our use of cookies and how you can manage your cookie settings, please see our Cookie Policy.  By continuing to use the website, you consent to our use of cookies.

Omdia view


NVIDIA is in the process of acquiring ARM from Softbank and plans to use its technology to advance the use of embedded artificial intelligence (AI) at the edge. NVIDIA stated that ARM will remain a UK-based company with its IP registered in the UK. NVIDIA also stated that it will set up an AI center of excellence in ARM’s Cambridge, UK campus. The focus of this acquisition is for NVIDIA to gain the ability to build low-power chips that will drive the embedded AI capability at the edge as smart devices, sensors, etc., become more common. NVIDIA will also help ARM accelerate its CPU roadmap in the data center to provide an alternative to legacy CPUs. Data center CPUs are a massive market that NVIDIA believes it can address, and it anticipates additional investment in ARM R&D to propel growth in new markets.

The edge and AI are both areas of growth

Figure 1 shows that, of the 11.9 million data center servers shipped in 2019, 2.4 million were deployed at the edge, up 8% from 2.2 million in 2018. The total 2019 revenue of all servers deployed at the edge (including tower, rack, blade, HCI, and open compute servers) totaled $14.7bn. Omdia expects the market for edge-deployed servers to continue to grow out to 2024 to enable low latency and secure the processing of large volumes of data as the count of IoT-connected devices grows and as enterprises increasingly adopt new software applications such as AI.

Figure 1: Omdia’s edge forecast

Omdia considers that the combined ecosystem of the two companies is a powerful statement about the growth of both edge and AI. With its range of solutions designed to support the building, training, and running of AI and machine learning (ML), NVIDIA is focused on the central development of AI/ML applications. ARM has shipped over 180 million units and is commonly found in small, low power devices, which is where AI edge applications will be executed. Therefore, Omdia can clearly see the value proposition for both companies, as developing AI/ML applications that can run efficiently in the location where they are needed, close to the user, requires the development environment and the runtime environment to be optimized accordingly.

As IoT devices continue to grow in number and complexity, they require a higher level of compute for management, control, support, and self-diagnosis. This is particularly the case when this is combined with the proliferation of real-time analytics and AI to improve business processes and develop new consumer services such as cloud gaming, AR/VR, 360° video, etc. A recent survey by Omdia on IoT (ICT Enterprise Insights 2019/20 – Global: IoT, Cloud, AI, and 5G) points to one important contributing factor. When asked about their biggest barrier to deployment for IoT projects, nearly half of all respondents (41.2%) pointed to the integration with existing IT infrastructure as either their first or second priority.

Omdia anticipates that investments in edge computing infrastructure such as NVIDIA’s acquisition of ARM demonstrates how the market will evolve. The edge computing marketplace spans a wide array of both hardware and software products, ranging from sensor to back-end services. Therefore, it makes sense that the market will look to build ecosystems and organizations that can offer seamless integration of both hardware and software for edge deployment and core development.



Roy Illsley, Chief Analyst, Cloud and Data Center Practice

askananaly[email protected]

Recommended Articles