Apple to run top AI model on Nvidia chips and Google Cloud - CNBC

Apple's decision to run top AI models on NVIDIA chips through Google Cloud represents a significant strategic move in the rapidly evolving AI landscape. This collaboration highlights the growing impor... Read original →

Apple’s reliance on NVIDIA GPUs via Google Cloud for its flagship AI models signals a strategic retreat from short-term in-house AI compute ambitions, opting instead for external high-performance infrastructure. This move amplifies demand for datacenter training chips, reinforcing NVIDIA’s dominance in the H100/B100 ecosystem and pressuring TSMC (Taiwan, China) to prioritize CoWoS advanced packaging capacity. From a compliance standpoint, the tripartite arrangement sidesteps immediate exposure to U.S. export controls on AI chips to China, yet deepens dependence on U.S.-centric cloud and silicon—a vulnerability in escalating geopolitical tensions. Competitively, Microsoft will likely accelerate AMD MI300X integration on Azure, while AWS may push Graviton+Trainium as a differentiated stack. Over the next 12–24 months, this partnership will cement ‘cloud-chip-model’ vertical integration as the industry norm, compelling device makers globally to reassess the economics and strategic viability of proprietary AI silicon.