Jassy Defends $200B AI Spend, Challenges Nvidia with Custom Chips
Cover image from businessinsider.com, which was analyzed for this article
In his annual shareholder letter, AWS CEO Andy Jassy defended Amazon's massive AI infrastructure spending and took aim at Nvidia's dominance while touting Amazon's custom AI chips. He outlined a vision for a leaner, AI-empowered workforce with smaller teams. The comments underscore Big Tech's AI arms race and supply chain diversification efforts.
PoliticalOS
Thursday, April 9, 2026 — Tech
Amazon is placing an enormous, long-term bet on custom AI chips and infrastructure to reduce reliance on Nvidia while reshaping its own workforce around smaller, AI-augmented teams. Jassy insists customer demand and efficiency gains justify the $200 billion spend despite bubble concerns. The decisive test will be whether Trainium, Leo satellites and related commitments deliver measurable revenue and performance gains over the next two years rather than remaining forward-looking assertions.
What outlets missed
Outlets underplayed the gap between Amazon's $20 billion chip run rate and Nvidia's $130 billion scale, making the competitive threat appear closer than current figures support. They also gave minimal attention to Amazon's satellite project history, including multiple delays and only about 30 prototype units in orbit versus Starlink's thousands of operational satellites serving paying customers. Internal dissent went largely unmentioned, including a December 2025 open letter signed by more than 1,000 Amazon employees criticizing the pace of AI investment for sidelining job impacts, worker well-being and climate considerations. Specific anecdotes, such as the exact six-engineer Bedrock team timeline, appear only in select coverage and lack independent corroboration beyond the letter itself. Few pieces examined whether the 'nearly sold out' claims for unreleased Trainium4 reflect binding contracts or softer reservations that could still evaporate.
Investors uneasy about sky-high valuations and talk of an AI bubble just heard from the leader of one of its biggest players. Amazon CEO Andy Jassy used his annual shareholder letter to justify another year of enormous spending while arguing that the company is wresting control of its AI future away from dominant suppliers.
The central tension is straightforward: virtually all AI training to date has run on Nvidia hardware, yet Jassy insists a shift is underway toward Amazon-designed silicon that delivers better price-performance for customers. According to the letter, capacity for the latest Trainium3 chips is already nearly sold out. The still-unreleased Trainium4, 18 months from availability, has also seen nearly all capacity reserved. Jassy pegged the internal chip business at a $20 billion annual revenue run rate. External sales at similar scale would push that figure near $50 billion. Nvidia, by comparison, recorded $130.5 billion in revenue for its fiscal year ended January 2026.