Morgan Stanley expects AI infrastructure spending to swell to $1.4 trillion by 2027, topping current market forecasts of about $1.2 trillion for next year, citing capacity constraints and rising capital plans at Amazon, Microsoft, Alphabet and Meta. Analyst Erik Woodring points to persistent demand outstripping supply across hyperscalers as a driver for elevated capex. The Motley Fool highlights Nvidia as a core beneficiary, citing dominance in AI training via GPUs and CUDA, plus moves into inference and CPUs, and says AMD could capture share in the faster-growing inference market through memory-rich chiplet designs and partnerships. The article also references acquisitions and partnerships—such as Nvidia’s work tied to Groq and AMD’s ties with Cerebras—as positioning both chipmakers to benefit from the anticipated spend.
Related articles:
CUDA Overview
AWS News Blog: AI/ML Infrastructure Updates
Microsoft Azure Blog: AI Infrastructure and Compute Updates
