CoreWeave, a provider of cloud services for GPU-powered apps, gets $50M

CoreWeave, a specialized cloud services provider for GPU-based workloads, said it has partnered with GPU giant Nvidia to enable companies to scale their infrastructure, cut costs, and improve efficiencies.

This rather new cloud services provider is offering on-demand computing resources across several verticals in the tech space and said it will continue to improve performance with help from a $50 million round of investment from Magnetar Capital.

Data scientists, machine learning engineers, and software engineers, to name a few roles, can utilize CoreWeave to deploy artificial intelligence applications and machine learning models. Furthermore, clients can manage their complex AI solutions, powered by GPUs, with the help of CoreWeave’s specialized DevOps expertise, the company said.

No more slow predictions from cloud services

A baseline problem in data science modeling from legacy cloud service providers is slow prediction. CoreWeave said it can improve prediction times by half by reducing inference latency, as well as servicing requests three times faster. As a result, the company estimated that cloud costs could be cut by 75%. These benefits can translate to a decrease in downtime, which can open up data scientists to allocate more time to other projects, ultimately allowing for a more comprehensive, data science-enabled company, CoreWeave suggested. Additionally, these cost savings can be put toward either hiring for more talent or investing in other resources instead.

“$50 million is a big number, and it allows us to accelerate our growth, fortify our infrastructure, and expand our footprint in ways previously unimaginable, despite being profitable since day one,” CoreWeave’s CEO, Michael Intrator, wrote in a blog post announcing the funding. “In 2022, we’ll be launching data centers in new regions, doubling down on our commitment to deliver the industry’s broadest range of high-end compute, and continuing to provide the world’s best infrastructure for on-demand, compute-intensive workloads.”