Source-led article
Amazon Eyes Direct Challenge to Nvidia in AI Chip Market with AWS Sales

Amazon Web Services (AWS) is reportedly in discussions to sell its internally developed AI chips, including its Trainium series, to other companies for deployment in their data centers. This strategic shift, initially hinted at by Amazon CEO Andy Jassy, signifies Amazon’s ambition to directly compete with established AI chip manufacturers like Nvidia in a market estimated to be worth $50 billion.
The move marks a significant evolution in AWS’s approach to its custom silicon. Traditionally, AWS has primarily used these chips within its own cloud infrastructure to power its AI services, offering customers access through its platform. Selling these chips directly would open up a new revenue stream and allow AWS to tap into the growing demand for specialized AI hardware beyond its cloud ecosystem.
Key facts
| Aspect | Detail |
|---|---|
| Company | Amazon Web Services (AWS) |
| Product | Custom AI chips (e.g., Trainium) |
| Strategy | Selling chips directly to third-party data centers |
| Market | Estimated $50 billion opportunity, targeting Nvidia’s dominance |
Expanding Beyond Internal Use
Amazon’s AI chief, Peter DeSantis, confirmed that AWS is in early-stage talks regarding the direct sale of its Trainium AI chips. While specific potential buyers were not disclosed, the initiative stems from CEO Andy Jassy’s earlier statements in his annual shareholder letter, where he highlighted the high demand for Amazon’s homegrown AI chips. Jassy estimated that if Amazon’s chip business operated as a standalone entity selling to both AWS and external parties, it could generate an annual run rate of approximately $50 billion.
This potential market entry represents a direct challenge to Nvidia, which currently holds a dominant position in the AI chip market. While Nvidia’s current revenue run rate significantly exceeds Amazon’s projected figure, a $50 billion competitor would still represent a substantial force, comparable to Intel’s annual revenues.
Implications for the AI Hardware Landscape
The decision to sell AI chips directly introduces several complexities for AWS. Historically, AWS has benefited from a “waterfall effect” where its chips power AI tokens, but also drive demand for a suite of supplementary services like storage, security, and networking within its cloud. Selling chips outright could alter this dynamic.
Another factor is the existing demand for AWS’s internal chip capacity. Jassy previously noted that current Trainium capacity, and even the future Trainium4, has sold out almost instantly. Direct sales to third parties would require AWS to significantly scale up manufacturing, potentially through partners like TSMC, without compromising its ability to serve its existing cloud customers. This could put AWS in direct competition with Nvidia for foundry capacity at a time when TSMC is a critical supplier for both.
For Indian businesses and AI developers, this development could mean increased options for high-performance AI hardware. As the IndiaAI Mission gains momentum and local startups push the boundaries of AI innovation, access to diverse and powerful AI chips becomes crucial. Amazon’s entry into the direct sales market could foster greater competition, potentially leading to more competitive pricing and wider availability of advanced AI accelerators for data centers and enterprises in India.
Source: TechCrunch AI – https://techcrunch.com/2026/06/18/amazon-hopes-to-challenge-nvidia-more-directly-by-selling-its-ai-chips/