1 month ago
33
A new group of cloud firms built for artificial intelligence workloads is drawing large investments and major enterprise clients. One example is a US$2 billion investment by Nvidia in Nebius, a company that runs large AI systems in the cloud.
The investment gives Nvidia an 8.3% stake in Nebius and signals rising demand for computing power from companies building or deploying AI systems. Reporting by Reuters says the deal reflects growing interest in a new category of providers often called “neocloud” companies. These firms focus almost entirely on high-performance infrastructure for AI model training and related workloads.
Traditional cloud providers host many kinds of software services. Neocloud firms take a different approach and focus on large clusters of graphics processors and networking systems used to train and run machine learning models. Demand for this type of infrastructure has grown as AI systems increase in size and complexity.
New cloud infrastructure
Nebius plans to expand its computing capacity in the coming years to support demand. The company intends to build more than five gigawatts of AI data-centre capacity by 2030, according to Reuters.
Facilities of this scale are designed to run large clusters of graphics processing units, which are used to train and operate AI models. Much of the demand for this capacity comes from large technology companies. Reuters said Nebius has secured major contracts to supply AI infrastructure to firms like Microsoft and Meta Platforms.
The agreements include a deal worth about US$17.4 billion with Microsoft and another valued at roughly US$3 billion with Meta, Reuters reported. The contracts show how large technology companies are expanding their access to computing power as AI workloads grow.
Training large AI systems requires vast computing capacity. Modern models often rely on thousands of graphics processors working together over long periods, and once deployed they must process large volumes of user requests in real time.
Nvidia expands its role in the cloud ecosystem
Nvidia’s investment fits into a broader effort by the chipmaker to expand its role in the AI computing stack. The company already supplies many of the graphics processors used in data centres that power AI systems. By investing in infrastructure providers, Nvidia can help increase the supply of computing environments built around its hardware. Reuters reported that the stake in Nebius is intended to support the expansion of AI cloud capacity as demand continues to grow.
The investment shows how the cloud market is splitting into new segments. For many years, most enterprise cloud workloads ran on a small group of large providers, including companies like Amazon through its AWS platform, Microsoft’s Azure service, and Google Cloud. AI workloads are now pushing that model in new directions. Many companies still rely on hyperscale cloud providers, while others are turning to specialist infrastructure firms that focus on GPU computing and high-performance networks.
Industry analysts often refer to this emerging group as “neocloud” providers. Companies like Nebius, CoreWeave, and Lambda are often placed in this category. Their services focus on AI model training and inference workloads, and many also support large-scale data processing.
Reshaped cloud supply
Many companies do not own the specialised hardware needed to run these systems at scale, so they rent computing capacity from cloud infrastructure providers. The model resembles earlier phases of cloud adoption, when businesses moved storage and applications to remote servers.
The difference today lies in the type of computing resources required. AI systems rely on specialised processors and fast data networks, and they require large data centres designed to support dense computing clusters. Building this infrastructure can cost billions of dollars. Accessing this capacity through external cloud providers is often more practical than building dedicated facilities.
A growing market for AI cloud capacity
The expansion of AI infrastructure is becoming one of the main drivers of cloud spending. Major technology companies have committed hundreds of billions of dollars to data centre construction and AI hardware purchases in recent years. Specialist providers are raising large sums to build facilities designed for AI workloads, sites housing clusters of GPUs and the cooling systems needed.
Nebius’s plan to build more than five gigawatts of data-centre capacity by 2030 shows the scale of these projects, which can support large computing campuses designed to train and run AI systems. Reuters said Nvidia’s US$2 billion investment will help Nebius expand its AI cloud infrastructure and meet rising demand from enterprise customers.
As companies deploy AI tools in products and internal systems, demand for this type of cloud infrastructure is likely to increase.
(Photo by Mackenzie Marco)
See also: Amazon plans huge AWS investment to meet AI cloud demand
Want to learn more about Cloud Computing from industry leaders? Check out Cyber Security & Cloud Expo taking place in Amsterdam, California, and London. This comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
CloudTech News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.