[ad_1]
CoreWeave, a specialised cloud supplier constructed for large-scale GPU-accelerated workloads, has secured $221 million in a Collection B spherical led by different asset supervisor Magnetar Capital. NVIDIA, together with former GitHub CEO Nat Friedman and former Apple government Daniel Gross, participated within the spherical.
Earlier than its concentrate on delivering cloud infrastructure, CoreWeave operated as an Ethereum mining firm. In accordance with a press launch, the corporate’s newest funding will allow it to proceed increasing its specialised cloud infrastructure that caters to compute-intensive duties reminiscent of synthetic intelligence, machine studying, visible results and rendering, batch processing, and pixel streaming to fulfill the rising calls for in generative AI know-how.
The brand new funds may also facilitate the enlargement of CoreWeave’s knowledge heart operations inside the US, together with the opening of two new facilities this 12 months. This can carry the whole variety of North American-based knowledge facilities operated by CoreWeave to 5.
“With the seemingly limitless boundaries of AI purposes and applied sciences, the demand for compute-intensive {hardware} and infrastructure is greater than it’s ever been,” stated Ernie Rogers, Magnetar’s chief working officer. “CoreWeave’s modern, agile and customizable product providing is well-situated to service this demand, and the corporate is consequently experiencing explosive development to help it.”
NVIDIA’s participation within the funding spherical additionally marks an enlargement of its collaboration with CoreWeave. On the NVIDIA GTC convention in March, NVIDIA unveiled its newest knowledge heart GPU, the NVIDIA H100 Tensor Core, and the NVIDIA HGX H100 platform.
CoreWeave additionally introduced on the convention that it had launched its HGX H100 clusters, that are presently utilized by purchasers like Anlatan, the creators of NovelAI. Together with HGX H100, CoreWeave supplies over 11 NVIDIA GPU SKUs which might be interconnected utilizing the NVIDIA Quantum InfiniBand in-network computing platform. These sources can be found to purchasers by way of reserved occasion contracts or on-demand.
Commenting on at present’s information, Manuvir Das, Vice President of Enterprise Computing at NVIDIA, stated: “AI has reached an inflection level, and we’re seeing accelerated curiosity in AI computing infrastructure from startups to main enterprises. CoreWeave’s technique of delivering accelerated computing infrastructure for generative AI, massive language fashions, and AI factories will assist carry the highest-performance, most energy-efficient computing platform to each trade.”
Learn extra:
[ad_2]
Source link
