Amazon Sells AI Chips Outside Its Own Cloud: A Strategic Game-Changer in the AI Hardware Market

2026-06-18T20:00:00 · Claude (Anthropic) · claude-sonnet-4-6

Amazon is making a major strategic move by selling its own AI chips outside the AWS cloud platform. This shift could significantly disrupt the chip market, currently dominated by NVIDIA, and offers businesses worldwide new possibilities for their AI infrastructure.

Amazon is significantly expanding its AI ambitions by selling its own AI chips outside its cloud infrastructure. This is a groundbreaking move for the tech giant, which until now has made its custom chips — such as the Trainium and Inferentia processors — exclusively available through Amazon Web Services (AWS). The new strategy opens the door to a direct confrontation with chip giant NVIDIA in the AI hardware market, with far-reaching consequences for businesses worldwide seeking affordable and powerful AI computing power.

Amazon's Own AI Chips: From Internal Use to the Open Market

Amazon has invested heavily in developing its own silicon over the past several years. The AWS Trainium chip is designed for training large AI models, while the AWS Inferentia chip is optimized for efficiently running AI inference — the process by which a trained model is deployed for real-world tasks. Both chips were developed by Amazon as a cost-effective alternative to NVIDIA's expensive GPUs.

Until now, access to these chips was exclusive to AWS customers: businesses wanting to benefit from Trainium or Inferentia had to use Amazon's cloud infrastructure. With this new step to offer these chips outside its own cloud, Amazon is fundamentally changing its position in the AI value chain and entering the market for standalone AI hardware.

Direct Competition with NVIDIA

Amazon's decision has far-reaching implications for the AI chip market. NVIDIA currently dominates this market with its H100 and H200 GPUs, which have become the gold standard for AI training and inference. Demand for NVIDIA chips has outstripped supply for years, driving extreme prices and long lead times for businesses looking to scale.

By offering its own chips on the open market, Amazon is positioning itself as a serious challenger to NVIDIA. Companies looking to invest in on-premises AI infrastructure — or those that, for security, compliance, or cost reasons, do not want to fully migrate to the cloud — now have a concrete alternative. This is particularly relevant for large enterprises, governments, and sectors such as finance and healthcare, where data privacy and digital sovereignty carry significant weight.

What Does This Mean for Businesses?

The availability of Amazon AI chips outside of AWS gives organizations greater flexibility in their AI strategy. Businesses no longer have to choose between the lock-in of a single cloud provider and the high cost of NVIDIA hardware. With Amazon's chips, they can:

• Expand their own data center infrastructure with powerful, specialized AI processors
• Reduce costs compared to equivalent NVIDIA solutions
• Maintain greater control over their AI workloads and sensitive business data

Amazon also benefits from the economies of scale it has built through years of internal use of these chips. The processors are optimized for popular AI frameworks such as PyTorch and TensorFlow, and are compatible with widely used model architectures — including those from Meta's LLaMA and Anthropic's Claude.

Amazon in the Broader AI Race

This move fits into Amazon's broader strategy to strengthen its position in the AI landscape. Amazon's AI chief has previously stated that he expects the first commercial quantum computers to become available within seven years — technology that could eventually revolutionize computing power for AI applications. Amazon is simultaneously investing heavily in large language models through its Bedrock platform and has formed strategic partnerships with companies including Anthropic.

By competing at the hardware level as well, Amazon closes the loop: from chips and cloud to AI models and services. This makes Amazon one of the most vertically integrated players in the AI ecosystem, comparable to how Apple combines hardware, software, and services within one tightly controlled ecosystem. To understand how we arrived at this moment, it is worth reading about the history of artificial intelligence.

Europe and Its Dependence on AI Hardware

Amazon's move comes at a time when Europe is grappling with its reliance on American and Asian chip manufacturers. The question of whether Europe risks being left behind in the so-called "AI war" highlights how urgent the need is for sovereign AI infrastructure and greater control over digital technology. Amazon's decision to sell chips outside its cloud may offer European businesses and governments more options, but it simultaneously deepens strategic dependence on large American tech giants — a paradox that policymakers in Brussels and beyond have yet to resolve.

Conclusion: Amazon Is Rewriting the Rules of the AI Chip Market

Amazon's decision to sell AI chips outside its own cloud is more than a commercial move — it is a strategic repositioning that could fundamentally reshape the balance of power in the AI hardware market. While NVIDIA works to maintain its dominant position and Intel and AMD continue to gain ground, Amazon joins the ranks of serious challengers with a unique advantage: years of internal validation on hyperscale AI workloads at one of the world's largest cloud providers.

For businesses and investors alike, this is a development to watch closely. The AI hardware market is at a tipping point, and Amazon is clearly playing the long game. Explore more AI news on stersoftware.com or dive deeper into the background via our knowledge base.

De Tijd

Source: De Tijd