<p>➀ Amazon is investing heavily in AI cluster construction with large-scale deployment of Hopper and Blackwell GPUs.</p><p>➁ AWS has allocated billions of dollars in capital expenditures to Trainium2 AI clusters.</p><p>➂ The article discusses the limitations of AWS's previous Trainium1 and Inferentia2 chips and the advancements made with Trainium2.</p>