Amazon plans to sell Trainium AI chips outside AWS in late 2027

2 weeks ago 50

Demand for Amazon's in-house AI accelerators has reportedly outrun what AWS can absorb, so the company is preparing to sell them elsewhere

Amazon.com Inc. (AMZN) price

Full chart and stats

Loading chart…

Market cap $2.7T (shares as of 2026-07-22)

P/E (trailing) 20.5 (EPS to 2026-06-30)

Annual revenue $716.9B USD (FY ended 2025-12-31) 10-K
Annual revenue growth +12.4% (FY ended 2025-12-31 vs FY ended 2024-12-31) 10-K
Annual net income 77,670,000,000 USD (FY ended 2025-12-31, attributable to the company) 10-K

Annual operating margin 11.2% (FY ended 2025-12-31)

Forward P/E Unavailable: no dated forecast source

Gloria daily close for session Oct 28, 2026. Market cap and trailing P/E use this close with the SEC share total and trailing diluted earnings, each dated next to its figure. Annual revenue, revenue growth, net income and margin show their own fiscal period and link the 10-K. Current context may postdate this article.

Amazon.com Inc. and its peers

Token vs underlying stock

Current comparison for AMZN. Historical observations may postdate this article.

AMZNx token unit vs AMZN share · USD

$254.56Oct 28 UTC

-6.7% over displayed period

● AMZNx token┄ AMZN stock: $254.60

Aug 5, 2026: token $272.84 at 20:00 UTC; stock close $272.65 Aug 6, 2026: token $272.13 at 20:00 UTC; stock close $272.26 Aug 7, 2026: token $274.76 at 20:00 UTC; stock close $274.48 Aug 10, 2026: token $277.96 at 20:00 UTC; stock close $278.09 Aug 11, 2026: token $272.21 at 20:00 UTC; stock close $272.27 Aug 12, 2026: token $267.48 at 20:00 UTC; stock close $267.28 Aug 13, 2026: token $265.07 at 20:00 UTC; stock close $265.13 Aug 14, 2026: token $262.81 at 20:00 UTC; stock close $262.65 Aug 17, 2026: token $261.07 at 20:00 UTC; stock close $261.31 Aug 18, 2026: token $259.40 at 20:00 UTC; stock close $259.45 Aug 19, 2026: token $265.80 at 20:00 UTC; stock close $265.84 Aug 20, 2026: token $259.95 at 20:00 UTC; stock close $260.11 Aug 21, 2026: token $258.73 at 20:00 UTC; stock close $258.63 Aug 24, 2026: token $262.28 at 20:00 UTC; stock close $262.07 Aug 25, 2026: token $261.06 at 20:00 UTC; stock close $261.06 Aug 26, 2026: token $260.32 at 20:00 UTC; stock close $260.28 Aug 27, 2026: token $256.30 at 20:00 UTC; stock close $256.26 Aug 28, 2026: token $266.44 at 20:00 UTC; stock close $266.43 Aug 31, 2026: token $259.63 at 20:00 UTC; stock close $259.77 Sep 1, 2026: token $254.64 at 20:00 UTC; stock close $254.92 Sep 2, 2026: token $254.81 at 20:00 UTC; stock close $254.98 Sep 3, 2026: token $258.83 at 20:00 UTC; stock close $258.90 Sep 4, 2026: token $258.34 at 20:00 UTC; stock close $258.51 Sep 8, 2026: token $256.81 at 20:00 UTC; stock close $256.97 Sep 9, 2026: token $252.24 at 20:00 UTC; stock close $252.40 Sep 10, 2026: token $251.61 at 20:00 UTC; stock close $251.89 Sep 11, 2026: token $257.13 at 20:00 UTC; stock close $256.78 Sep 14, 2026: token $253.68 at 20:00 UTC; stock close $253.54 Sep 15, 2026: token $248.35 at 20:00 UTC; stock close $248.42 Sep 16, 2026: token $246.01 at 20:00 UTC; stock close $245.96 Sep 17, 2026: token $251.35 at 20:00 UTC; stock close $251.19 Sep 18, 2026: token $253.69 at 20:00 UTC; stock close $253.71 Sep 21, 2026: token $258.31 at 20:00 UTC; stock close $258.45 Sep 22, 2026: token $255.06 at 20:00 UTC; stock close $254.98 Sep 23, 2026: token $249.39 at 20:00 UTC; stock close $249.27 Sep 24, 2026: token $249.48 at 20:00 UTC; stock close $249.38 Sep 25, 2026: token $249.71 at 20:00 UTC; stock close $249.67 Sep 28, 2026: token $246.30 at 20:00 UTC; stock close $246.15 Sep 29, 2026: token $246.66 at 20:00 UTC; stock close $246.67 Sep 30, 2026: token $249.53 at 20:00 UTC; stock close $249.15 Oct 1, 2026: token $248.17 at 20:00 UTC; stock close $248.23 Oct 2, 2026: token $251.37 at 20:00 UTC; stock close $251.52 Oct 5, 2026: token $251.43 at 20:00 UTC; stock close $251.40 Oct 6, 2026: token $256.28 at 20:00 UTC; stock close $256.29 Oct 7, 2026: token $259.85 at 20:00 UTC; stock close $259.92 Oct 8, 2026: token $254.31 at 20:00 UTC; stock close $254.06 Oct 9, 2026: token $262.35 at 20:00 UTC; stock close $262.43 Oct 12, 2026: token $259.60 at 20:00 UTC; stock close $259.53 Oct 13, 2026: token $258.72 at 20:00 UTC; stock close $258.58 Oct 14, 2026: token $253.89 at 20:00 UTC; stock close $254.08 Oct 15, 2026: token $254.89 at 20:00 UTC; stock close $254.76 Oct 16, 2026: token $250.16 at 20:00 UTC; stock close $250.00 Oct 19, 2026: token $262.52 at 20:00 UTC; stock close $262.45 Oct 20, 2026: token $262.57 at 20:00 UTC; stock close $262.63 Oct 21, 2026: token $262.57 at 20:00 UTC; stock close $262.37 Oct 22, 2026: token $250.30 at 20:00 UTC; stock close $250.38 Oct 23, 2026: token $248.07 at 20:00 UTC; stock close $248.00 Oct 26, 2026: token $248.95 at 20:00 UTC; stock close $249.09 Oct 27, 2026: token $253.41 at 20:00 UTC; stock close $253.10 Oct 28, 2026: token $254.56 at 20:00 UTC; stock close $254.60 Aug 5$245.96 to $278.09Oct 28

Amazon plans to start offering its Trainium AI chips outside of Amazon Web Services in late 2027. The company says the move responds to customer demand.

Trainium is Amazon’s custom silicon for training and running AI models. Think of it as Amazon’s house-brand alternative to the GPUs most AI labs buy from Nvidia.

Until now, the only way to use those chips was to rent them as part of AWS’s cloud. Selling them outside AWS would let other data centers run Trainium hardware directly.

Peter DeSantis, who leads AI at AWS, has confirmed that Amazon is in talks with potential buyers. Those discussions were taking place as of mid-2026.

The timeline stays loose. Late 2027 is the window for meaningful external availability. Amazon has not disclosed specific timelines or pricing for these sales.

The reason is simple: there isn’t enough Trainium to go around. Capacity through AWS is reported to be mostly sold out for 2026. It is reportedly largely committed for 2027 and parts of 2028 as well.

The numbers behind the demand

The financial commitments are large. More than $225 billion in revenue commitments tied to Trainium have been made, involving major AI firms such as OpenAI and Anthropic.

Both labs are committed to using Trainium. Each plans to ramp up its capacity significantly in the coming years.

Amazon CEO Andy Jassy has suggested the broader chip segment could reach more than $20 billion in annual revenue by 2026. That segment covers both Trainium and Graviton, Amazon’s general-purpose processor line.

Trainium3 delivers a 30-40% improvement in price-performance over Trainium2. Trainium2 was itself approximately 30% better than existing GPUs on that measure.

Background: the software problem

Fast chips are only half the battle in AI hardware. The other half is software, and that is where Nvidia has built its moat.

Nvidia’s CUDA platform is the toolkit most AI developers grew up on. Years of code, libraries and engineering habits are tied to it. Building a comparable ecosystem around a rival chip is a formidable challenge, regardless of how good the silicon is.

Amazon is trying to lower that barrier. In October 2026, it launched TorchNeuron and open-sourced it. The tool lets developers move PyTorch code onto Trainium chips with minor adjustments.

What this means

For Amazon, external sales would add a new revenue stream on top of its cloud rental business. Today, Trainium income runs through AWS. Selling hardware to outside data centers would let Amazon earn from demand it currently cannot serve.

For investors, the picture comes with real upside and clear unknowns. The $225 billion-plus in commitments and Jassy’s $20 billion-plus revenue outlook point to a chip business with serious scale.

No pricing has been disclosed, and pricing will decide whether Trainium can win buyers away from Nvidia. The software ecosystem remains a hurdle, since TorchNeuron is still new and CUDA’s head start is measured in years.

There is also a supply question. If Trainium is already reportedly committed into 2028 through AWS alone, external buyers will be competing for the same limited output.

Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy.

Read Entire Article