CoreWeave ‘came out of nowhere.’ Now it’s poised to make billions off AI with its GPU cloud

A few months ago, few had heard of CoreWeave, a cloud startup specializing in GPU-accelerated workloads, according to Brannin McBee, the company’s co-founder and chief strategy officer. Now, CoreWeave is poised to make billions off of the generative AI boom with its GPU cloud.

“We’ve come out of nowhere,” he told VentureBeat in an interview last week. With over $400 million in new funding; a new $1.6 billion data center in Plano, Texas; and the world’s fastest AI supercomputer built in partnership with Nvidia unveiled last month, the company’s fortunes have shifted dramatically — thanks, in no small part, to Nvidia. 

CoreWeave: from crypto mining to GPU acceleration at scale

CoreWeave was founded in 2017 by three commodities traders who turned their cryptocurrency mining hobby into an Ethereum mining company, using GPUs to verify blockchain transactions — doing business out of a New Jersey data center. By 2019, the founders had pivoted — fortuitously, in hindsight — to building a specialized cloud infrastructure spanning seven facilities that offered GPU acceleration at scale. 

Suddenly, everyone was talking about CoreWeave, which led to an investment “tide shift” in March of this year, said McBee. “People were still able to access GPUs last year, but when it became extremely tight, all of a sudden it was like, where do we get these things?” he explained. AI companies that were using CoreWeave spread the word to VCs, he added, who suddenly saw a gold mine: “They said, ‘Why aren’t we speaking to these guys’?”

That led to a massive $221 million Series B funding round in April, which included an investment from Nvidia, and one month later, CoreWeave secured another $200 million. 

McBee said CoreWeave did $30 million in revenue last year, will score $500 million this year and has nearly $2 billion already contracted for next year. CNBC reported in June that Microsoft “has agreed to spend potentially billions of dollars over multiple years on cloud computing infrastructure from startup CoreWeave.”

“It’s happening very, very quickly,” he said. “We have a massive backlog of client demand we’re trying to build for. We’re also building at 12 different data centers right now. I’m engaged in something like one of the largest builds of this infrastructure on the planet today, at a company that you had never heard of three months ago.”

Nvidia has diverted latest AI server chips to CoreWeave

In addition to being in the right place with the right technology at the right time, CoreWeave has also benefitted significantly from Nvidia’s strategy to stay dominant in the AI space.

Nvidia has allotted a generous number of its latest AI server chips to CoreWeave and away from top cloud providers like AWS, even though supply is tight. That’s because those companies are developing their own AI chips in an attempt to reduce their reliance on Nvidia.

“It’s certainly isn’t a disadvantage to not be building our own chips,” McBee admitted. “I would imagine that that certainly helps us in our constant effort to get more GPUs from Nvidia at the expense of our peers.”

But while having Nvidia in their corner is “excellent” for CoreWeave, ultimately, McBee said, over time there will a matrix of different pieces and types of infrastructure that support different types of AI models. However, he believes GPUs will remain the infrastructure that supports the most cutting-edge, most compute-intensive models that get developed — and that it will take at least another two years, if not three, for the GPU supply shortage to begin to alleviate.

CoreWeave clients include Inflection AI

For now, top AI companies like Inflection AI, which recently announced an eye-popping $1.3 billion funding round to build a massive GPU cluster, are using CoreWeave to build it.

“They called us and said, ‘Guys, we need you to build one of the most high-performance supercomputers on the planet to support our AI company,’” McBee said. “They call us and they say, ‘This is what we’re looking for, can you do it?’ — and I think we have two of the top five supercomputers that we are building right now in terms of FLOPS.”

For a client like Inflection, he explained, CoreWeave comes up with a timeline for the large build, and then explains to Nvidia what they are doing. “They say, ‘We’ll support you with engineering, marketing, infrastructure, allocation, whatever you need to get this done,’” he said. “Then we go execute, and that’s also where our background comes in — we’ve consistently executed and we’ve built a differentiated product in the market.”

The future looks as bright as the GPU gold rush for CoreWeave

Now that it has become a household name in AI, CoreWeave is making plans to establish itself in industries such as life sciences — for areas like drug discovery, protein folding simulations, molecular discovery, and genetics testing. “That all requires the type of compute that we’re operating and they need to be able to access it in massive scale,” he explained.

But at the moment, McBee sounds a bit like a kid in a GPU candy store when it comes to CoreWeave’s efforts to power the generative AI boom.

“Walking into the data centers is the most amazing thing,” he said. “I’ve been building computers since I was a kid. So it’s just fun to be able to be in a business where you’re actually able to do this and scale it and power some of the coolest companies on the planet right now.”