AI hyperscalers plan to lift data center capex from $650 billion in 2026 to $1 trillion in 2027, priming Nvidia’s Vera Rubin processors for strong revenue gains. Apple will run its revamped Siri on Nvidia Blackwell B200 GPUs via Google Cloud while Cerebras raised $5.5 billion in its IPO to challenge Nvidia’s share.
Global AI hyperscalers are set to boost data center capital expenditures from $650 billion in 2026 to $1 trillion in 2027. This surge underscores the growing investment in AI infrastructure and positions Nvidia to capture a significant share of procurement for next-generation GPUs.
Nvidia’s upcoming Vera Rubin architecture promises higher throughput and energy efficiency for large-scale AI training and inference workloads. Elevated pricing and broad hyperscaler demand are expected to drive meaningful revenue gains for Nvidia in fiscal 2027.
Apple plans to deploy Nvidia Blackwell B200 GPUs via Google Cloud to power its revamped Siri voice assistant, leveraging Nvidia’s confidential compute feature to maintain end-to-end data encryption. This marks a deeper reliance on external AI infrastructure to enhance Siri’s capabilities.
Cerebras Systems raised $5.5 billion in its recent IPO, securing partnerships with AWS and OpenAI to deploy its wafer-scale processors. While its unique design promises faster inference, customers must adopt new hardware platforms, presenting both opportunity and friction in competing with Nvidia’s ecosystem.
Finance
