The Race to Beat Nvidia: Does Google or Amazon Have the Better In-House Silicon

Google and Amazon each poured billions into custom silicon and talked up their chips on the same earnings day, but their strategies point in opposite directions. Only one looks like the smarter bet right now, and it may not be…

Published July 17, 2026, 2:01pm ET · 2 min read

A graphic featuring the Nvidia logo on the left and the Google logo on the right, separated by a "VS" icon with a yellow lightning bolt, against a dark hexagonal grid background that shifts to green on the right. Below the logos, text reads "Nvidia Blackwell vs. Google TPU". At the top, a banner declares "THE AI SHOWDOWN YOU CAN'T IGNORE". A man with a surprised expression is visible in the bottom right corner.
This graphic captures the intense rivalry between Nvidia's Blackwell architecture and Google's TPU in the rapidly evolving AI silicon market. It visually represents the competitive landscape discussed in the article regarding AI hardware. © 24/7 Wall St

Google (NASDAQ: GOOGL | GOOGL Price Prediction) and Amazon (NASDAQ: AMZN) both reported Q1 FY2026 results on April 29, 2026, and each spent much of the call discussing custom silicon. With NVIDIA (NASDAQ: NVDA) still setting the pace, Google’s TPU roadmap and Amazon’s Trainium franchise represent the two most credible in-house alternatives in the cloud.

TPU Powers Gemini. Trainium Powers Everyone Else

Google’s silicon story is vertical. Sundar Pichai told investors that “we own frontier models and own the silicon”, and the payoff showed up in Cloud, which grew 63% to $20 billion with backlog nearly doubling to $462 billion. The new 8th-generation TPU 8t offers 3x the processing power of Ironwood, and Gemini now handles more than 16 billion tokens per minute via direct API. Pichai acknowledged “we are compute constrained in the near term”.

Amazon is selling picks and shovels. Andy Jassy said the chips business topped a $20 billion annual revenue run rate, growing triple digits year over year, with total Trainium commitments over $225 billion. AWS grew 28%, its fastest pace in 15 quarters, aided by multi-gigawatt training deals from Anthropic and OpenAI. Trainium2 is 30% cheaper per unit of performance than comparable GPUs. That dents NVIDIA’s pricing power.

Internal Moat vs Merchant Chip

Lens Google Amazon
Silicon Strategy TPU tied to Gemini stack Trainium sold to rival labs
2026 CapEx Guide $180B to $190B ~$200B
Anchor Customer Google itself Anthropic, OpenAI, Meta (NASDAQ: META)
YTD Stock +18.66% +10.46%

The philosophical split is clear. Google treats TPU as an internal weapon that occasionally ships to “capital markets” and select frontier AI labs. Amazon treats Trainium as a merchant product, with Jassy hinting “we are going to sell racks in the next couple of years”. One approach compounds margin. The other compounds distribution.

The Next Test Is 2027 Deliveries

I will watch whether Google converts that $462 billion backlog into shipped TPU capacity, since the CFO said the majority of TPU hardware revenue lands in 2027. For Amazon, the pressure point is Trainium3 supply, which Jassy called “nearly fully subscribed”. If memory costs keep skyrocketing, gross margins on custom silicon could compress before scale kicks in.

Why I Lean Toward Google Today and Amazon by 2027

Google screens as the cleaner near-term setup on the data. TPU v7 Ironwood and Google’s software ecosystem run frontier models at a real total-cost-of-ownership advantage, and a forward P/E around 25 is reasonable for 22% revenue growth. That mix is hard to fault.

Amazon’s thesis rewards a longer time horizon. I do not love the forward P/E near 29 today, but Trainium’s cadence and Amazon’s distribution muscle should make it the dominant non-NVIDIA hardware standard of the next cloud era. Google offers the moat visible today. Amazon’s moat is being poured in concrete for tomorrow.

Contact [email protected] for any questions or corrections.

Alex Sirois

Alex Sirois is a financial writer with experience spanning both retail and institutional investing. He has written for InvestorPlace and held roles at BNY Mellon and Bernstein, giving him a perspective that bridges Main Street portfolios and Wall Street analysis.

Alex holds an MBA from George Washington University and has built his career across multiple industries, including e-commerce, education, and translation — a breadth of experience that informs how he breaks down complex financial topics for everyday investors. His writing is conversational, actionable, and grounded in long-term, buy-and-hold investing principles.

At 247 Wall St., Alex focuses on delivering analysis that is both accessible and useful, with a clear emphasis on helping readers make more informed decisions with their money.

All articles →