NVIDIA's CEO Jensen Huang has doubled down on the company's commitments to CoreWeave, as the neocloud will be the first to leverage Vera CPUs as a standalone offering.
CoreWeave Will Offer Vera CPUs as "Standalone" Option, Indicating NVIDIA Is All-In Towards Agentic AI
It appears that NVIDIA is back in the capital markets after scoring huge deals with the likes of Groq and Intel, and this time, the company has committed an additional $2 billion to the neocloud CoreWeave, claiming it will complement "long-standing" relations. According to the company's official blog post, NVIDIA will purchase Class A common stock at $87.20 per share, helping CoreWeave achieve its ambition to build 5 gigawatts of AI factories by 2030. While talking with Bloomberg's Ed Ludlow, NVIDIA's CEO revealed that CoreWeave will be the first to have access to Vera CPUs under the new deal.
For the very first time, we're going to be offering Vera CPUs. Vera is such an incredible CPU. We're going to offer Vera CPUs as a standalone part of the infrastructure. And so not only, not only can you run your computing stack on NVIDIA GPUs, you can now also run your computing stack, wherever their CPU workload, run on Nvidia CPUs...
Vera is completely revolutionary...CoreWeave is going to have to race if CoreWeave's going to be the first to stand up Vera CPUs. We haven't announced any of our CPU design wins, but there are going to be many.
- NVIDIA's Jensen Huang via Ed Ludlow
Well, it appears that from this particular deal, we have two narratives in place. The first is that NVIDIA has recognized that server CPUs are becoming another major bottleneck in the AI supply chain, and that, with agentic AI applications ramping up, a viable platform is more necessary than ever. Secondly, by offering Vera CPUs as a "standalone" offering, NVIDIA aims to provide customers a workaround for those willing to consider high-end CPU capabilities, and it would be a much cheaper option compared to getting the entire rack-scale solution.
Vera CPUs are set to be among the most capable CPU offerings from NVIDIA, offering a massive bump over Grace Blackwell. You are looking at next-gen custom ARM architecture codenamed Olympus, which packs 88 cores, 176 threads (with NVIDIA Spatial Multi-Threading), a 1.8 TB/s NVLink-C2C coherent memory interconnect, 1.5 TB of system memory (3x Grace), 1.2 TB/s of memory bandwidth with SOCAMM LPDDR5X, and rack-scale confidential compute.
This direction clearly indicates that when it comes to the server CPU ecosystem, NVIDIA is looking to go "heavy", and interestingly, when Jensen mentions "CPU design wins", he might also be pointing towards the upcoming ARM-based N1/N1X SoCs, which are upcoming consumer offerings dedicated towards AI PC workloads.
About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.
Follow Wccftech on Google to get more of our news coverage in your feeds.
