"Nebius is a 'neocloud' built on deep expertise in hardware and software development. It builds large-scale GPU clusters and on top of that operates an AI-native cloud computing platform that gives AI innovators everything they need to build, train and run models and apps — all in one place.
Today, about half of all AI chips are used internally by major model developers — OpenAI, Microsoft, Google, and Meta — to build AI models and run inference, as they race to be the first to achieve Artificial General Intelligence.
Another 10-20% of compute capacity is available to the public market through three or four hyperscalers, such as Amazon and Oracle. The remaining 30-40% is distributed among various alternative players. While there are dozens of these companies, only a few have both the technology and capital to build large-scale infrastructure and sell tens of thousands of GPUs on the open market. This year, hundreds of thousands of GPUs will be sold.
In this space, we are one of the largest providers of alternative AI infrastructure and among the biggest players outside the U.S. Other companies in this category include CoreWeave, Lambda Labs, Together(.)ai, and Deep Infra. Competition is always good. Today, no one can build fast enough — the tech giants can barely meet their own demand.
We are one of the few alternatives capable of serving the core needs of major players, supplying GPUs to startups and corporate clients looking to purchase infrastructure, and supporting smaller clients who use models deployed by us. We are giving customers the freedom to choose."
Will Coreweave impact its business?
Check out the article I'm about to post! :)
Coul you tell me one thing that sums up the whole business?
Maybe this helps to understand its core business:
"Nebius is a 'neocloud' built on deep expertise in hardware and software development. It builds large-scale GPU clusters and on top of that operates an AI-native cloud computing platform that gives AI innovators everything they need to build, train and run models and apps — all in one place.
Today, about half of all AI chips are used internally by major model developers — OpenAI, Microsoft, Google, and Meta — to build AI models and run inference, as they race to be the first to achieve Artificial General Intelligence.
Another 10-20% of compute capacity is available to the public market through three or four hyperscalers, such as Amazon and Oracle. The remaining 30-40% is distributed among various alternative players. While there are dozens of these companies, only a few have both the technology and capital to build large-scale infrastructure and sell tens of thousands of GPUs on the open market. This year, hundreds of thousands of GPUs will be sold.
In this space, we are one of the largest providers of alternative AI infrastructure and among the biggest players outside the U.S. Other companies in this category include CoreWeave, Lambda Labs, Together(.)ai, and Deep Infra. Competition is always good. Today, no one can build fast enough — the tech giants can barely meet their own demand.
We are one of the few alternatives capable of serving the core needs of major players, supplying GPUs to startups and corporate clients looking to purchase infrastructure, and supporting smaller clients who use models deployed by us. We are giving customers the freedom to choose."
— Arkady Volozh, NBIS' Founder & CEO
Really useful, thanks! I think you’ve just convinced me to buy into their business haha.