Amazon Wants to Sell Its Own AI Chips

2026-06-18 · Source: AutoGPT · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Novice, quick

Summary

Amazon Web Services (AWS) is exploring the direct sale of its proprietary AI chips, specifically Trainium, to external companies for their own data centers, a move announced by AWS AI chief Peter DeSantis on June 18, 2026. This initiative, hinted at in Amazon CEO Andy Jassy's April 2026 shareholder letter, suggests Amazon's chip business could generate approximately \$50 billion annually if it operated independently. This strategy aims to challenge Nvidia, which currently holds a \$326 billion annual revenue run rate in the AI chip market. However, selling chips externally means AWS would forgo additional revenue from its cloud ecosystem services like storage and networking. Furthermore, existing Trainium chip capacity is already sold out, with the next version, Trainium4, also pre-sold over a year in advance, posing significant supply chain challenges, particularly in securing manufacturing capacity from TSMC, where Nvidia is the largest customer.

Key takeaway

For Directors of AI/ML evaluating future infrastructure investments, Amazon's move to sell Trainium chips directly signals a potential shift in AI hardware sourcing. You should assess how this new option could diversify your supply chain beyond Nvidia, potentially mitigating vendor lock-in and optimizing costs. Be aware of initial supply constraints and the trade-offs between cloud-integrated services and on-premise hardware control.

Key insights

Amazon's direct AI chip sales could disrupt Nvidia's dominance but face complex revenue and supply chain trade-offs.

Principles

• Cloud providers seek hardware independence.
• Ecosystem revenue complicates hardware sales.
• Supply chain capacity is a critical bottleneck.

In practice

• Evaluate proprietary chips for data centers.
• Assess cloud vs. on-premise AI hardware costs.
• Monitor TSMC capacity allocation shifts.

Topics

• AI Chips
• Amazon Web Services
• NVIDIA
• Trainium
• TSMC
• Data Center Infrastructure

Best for: CTO, AI Architect, MLOps Engineer, Director of AI/ML, VP of Engineering/Data, Investor

Related on AIssential

• Amazon hopes to challenge Nvidia more directly by selling its AI chips — Amazon is considering selling its Trainium AI chips to third parties, aiming to capture a \$50 billion market and directly challenge Nvidia.
• Amazon in Talks to Sell Custom AI Chips in Bid to Undercut Nvidia — Amazon aims to sell custom AI chips, directly challenging Nvidia's market dominance.
• Why Everyone Is Moving Away from NVIDIA — Hyperscalers are vertically integrating AI infrastructure, developing custom silicon and energy solutions to overcome GPU bottlenecks.
• Google to Release New AI Chips, Challenging Nvidia — Google is specializing its TPU program for AI inference to meet surging demand and maintain its competitive edge.
• Anthropic takes $5B from Amazon and pledges $100B in cloud spending in return — Major AI developers are securing massive compute capacity through cloud infrastructure deals with tech giants.
• 🔮 Exponential View #570: Inside Jensen Huang’s worldview — Nvidia's market dominance stems from strategic supply chain orchestration and a deeply embedded software ecosystem.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AutoGPT.