Amazon’s proprietary Trainium and Inferentia chips have reached nearly full capacity utilization across its key data centers, driven by strong enterprise adoption of AI training and inference workloads. The sell-out highlights AWS’s ability to scale proprietary silicon but also points to supply constraints as demand for cost-efficient AI hardware climbs.

Nvidia continues to deliver robust year-over-year Data Center revenue growth fueled by widespread GPU demand for AI applications. While AWS’s chip sell-out underscores competitive dynamics, Nvidia’s accelerating top-line performance suggests it maintains an edge in performance, software ecosystem, and customer preference.