Google teams up with CoreWeave to secure NVIDIA's computing power
Google is in talks with CoreWeave to rent data center servers equipped with NVIDIA's latest GPUs, demonstrating its thirst for computing power in the AI race. Google is also seeking backup GPU supply options from CoreWeave's competitors, indicating that it faces resource constraints in acquiring high-performance computing chips. Additionally, Google plans to lease data center space from CoreWeave to deploy its self-developed TPUs, showcasing the strategic significance of its computing power layout. CoreWeave's stock price surged by 16%, reflecting strong market expectations for computing power demand
According to informed sources, the Zhitong Finance APP has learned that Google (GOOGL.US) is in deep discussions with the computing infrastructure newcomer CoreWeave (CRWV.US), planning to rent data center servers equipped with NVIDIA's (NVDA.US) latest Blackwell architecture GPUs. Reports indicate that the two parties are close to reaching a final agreement. Although Google is already one of NVIDIA's largest purchasers, the move to acquire more cutting-edge computing power through third-party channels reflects the tech giant's extreme thirst for computing resources in the AI race.
Even more intriguing is that Google is simultaneously engaging with CoreWeave's competitors to seek backup GPU supply options. This "dual-track strategy" reveals a key signal: even as the world's largest cloud computing service provider, Google still faces resource constraints in acquiring high-performance computing chips and has to resort to unconventional means to break through computing power bottlenecks.
The cooperation between the two parties goes far beyond this. According to in-depth reports, Google is also negotiating another important collaboration with CoreWeave: leasing its data center space to deploy self-developed TPUs (Tensor Processing Units). This "computing power hybrid" model carries significant strategic implications—leveraging NVIDIA GPUs' general computing capabilities while deploying its own TPUs to enhance AI model training efficiency, showcasing the tech giant's foresight in computing power layout.
In response, the capital market reacted swiftly. CoreWeave's stock price surged 16% at Wednesday's close, marking the largest recent increase, reflecting strong investor expectations for an explosion in computing power demand. Google's and NVIDIA's stock prices remained relatively stable, but underlying currents are stirring: NVIDIA's position as a core supplier of computing power is becoming increasingly solid, while Google's strategic intent to strengthen its AI competitiveness through diversified cooperation has become clear
