Report: CoreWeave Wins Miscrosoft Deal for GPU Cloud Services Worth Billions

Print Friendly, PDF & Email

Suddenly, GPU cloud provider CoreWeave is one of the hot tech companies of the week and likely longer. No sooner did the Roseland, NY, startup announce on Wednesday it had secured $200 million in a series B venture round extension than the news broke the next day that CoreWeave has signed a deal with Microsoft that may be worth billions of dollars across multiple years, according to a CNBC story.

Attribute all of this, of course, to the generative AI boom kicked off last November by the launch of ChatGPT from OpenAI, the recipient of $13 billion in investment funds from Microsoft. Now tech vendors and companies are scrambling to leverage large language models, the training of which – a highly compute-intensive process – is powered by GPUs. And that leads to GPU market leader NVIDIA, a $100 million backer of, yes, CoreWeave.

In fact, after Wednesday’s venture round announcement (bringing total VC investment in CoreWeave to $421 million), one online wag wondered on Twitter how much of the $200 million in new venture capital will wind up in the NVIDIA’s coffers. With generative AI – and therefore GPUs – all the rage, NVIDIA share prices have gone up 170 percent since the beginning of the year. “The company’s market cap briefly topped $1 trillion for the first time this week after it issued a forecast for the July quarter that was over 50 percent higher than Wall Street estimates,” CNBC reported.

Along with NVIDIA, so often in the right place at the right time, it’s clear that CoreWeave’s ship has come in. As company CEO and co-founder Michael Intrator said in the company’s Wednesday venture announcement, “In my 25-year career, I’ve never been a part of a company that’s growing like this. It’s an incredible moment in time. From a demand standpoint, revenue and client scale, the rise has been exponential….”

CoreWeave CEO Michael Intrator

CoreWeave, founded in 2017 and with 160 employees, describes itself as offering a specialized cloud infrastructure with a massive scale of GPU compute resources to train, serve inference and fine-tune models. “CoreWeave builds cloud solutions for compute-intensive use cases — VFX and rendering, machine learning and AI, batch processing and pixel streaming — that are up to 35 times faster and 80 percent less expensive than the large, generalized public clouds,” the company states, adding that “CoreWeave’s serverless orchestration layer makes it an attractive option for AI startups rushing to get a product to market, as teams can spin up workloads in as little as five seconds with increased portability, less overhead and less management complexity.”

According to the CNBC article, “With so much demand for its infrastructure, Microsoft needs additional ways to tap NVIDIA’s GPUs.”