Amazon is reportedly in discussions to sell its custom-designed AI chips — known as Trainium — to companies outside of its AWS cloud division, a move that would position the e-commerce giant as a direct competitor to Nvidia in the lucrative AI hardware market.

The news, first reported by Bloomberg and confirmed by multiple sources including TechCrunch and Yahoo Finance, represents a significant strategic shift for Amazon. Until now, Trainium chips have been available exclusively to AWS customers who rent computing power on Amazon's cloud platform. For more context on this story, see our ongoing latest AI developments.

What We Know

According to Bloomberg's reporting, Amazon is exploring ways to make its Trainium chips available to external buyers — potentially including server manufacturers and data center operators who currently rely heavily on Nvidia GPUs. The move would allow companies to purchase Amazon-designed chips directly rather than accessing them only through AWS.

The strategy is part of a broader effort to undercut Nvidia's pricing power in the AI chip market, where Nvidia currently holds an estimated 80% market share.

Why It Matters

The AI chip market has become one of the most valuable sectors in technology. Nvidia's dominance has driven its market capitalization past $3 trillion, but competitors are increasingly aggressive:

  • AMD has expanded its Instinct MI300 series for AI workloads
  • Google continues developing its Tensor Processing Units (TPUs)
  • Microsoft has begun designing custom Maia AI chips
  • Meta is building its own MTIA inference chips

Amazon's entry into direct chip sales could disrupt the market dynamics by offering a credible alternative at potentially lower prices.

The Trainium Advantage

Amazon's Trainium chips are designed specifically for training large language models and other AI workloads. AWS has reported that some AI startups are already choosing Trainium over Nvidia GPUs for cost reasons.

According to Amazon's own reporting, startups building AI applications beyond chatbots — including those working on drug discovery, financial modeling, and autonomous systems — have been early adopters of the chip line.

What Analysts Say

Industry analysts are divided on the potential impact:

Some view the move as a natural evolution of Amazon's chip strategy, noting that the company has invested billions in its semiconductor division through acquisitions like Annapurna Labs.

Others caution that Nvidia's software ecosystem — particularly CUDA, its proprietary programming platform — remains a significant moat. Developers have spent years building tools and libraries around CUDA, and switching costs are high.

Nvidia's Response

Nvidia has not publicly commented on Amazon's plans. However, the company has been accelerating its own roadmap, with next-generation Blackwell and Rubin chip architectures in development.

Nvidia has also been expanding beyond pure hardware, offering AI software platforms, networking equipment, and full data-center reference designs to maintain its competitive position.

Looking Ahead

If Amazon proceeds with selling chips externally, it could reshape the competitive landscape of AI hardware. The move would signal that the company sees chip design as a core business rather than just a way to differentiate AWS.

For AI companies and researchers, more competition in the chip market likely means lower prices and more options — a win for the broader AI ecosystem.

---

Stay Ahead of AI

Get the latest AI news, analysis, and breakthroughs — all in one place.

Read more AI news →