Alphabet is planning a new AI chip. Here's why it matters ahead of earnings

Alphabet (NASDAQ: GOOGL) is in the midst of one of its most consequential weeks of the year, and it started with a bang. On Monday, July 20, it was reported that the company is developing a new AI chip, internally dubbed Frozen v2, designed to run its Gemini models dramatically more efficiently. The stock closed over 1% higher on the news, and is now up close to 13% year to date. With Q2 earnings due Wednesday, July 22, after the close, the timing of the report could hardly be more relevant.

What Alphabet's Frozen v2 AI Chip Could Mean for Gemini

The chip represents a genuinely different approach to AI silicon. Rather than building general-purpose processors that load models into memory and shuffle data back and forth, Frozen v2 would permanently embed parts of Gemini's architecture directly into the silicon itself. That reduces the number of calculations and the amount of data movement required to answer a query. Google engineers reportedly estimate the chip could serve 6 to 10 times as many tokens per unit of power as the company's newest TPUs, with deployment targeted as soon as 2028.

There are trade-offs, though. Hardwiring the model into the chip means it only works with future Gemini versions if Google keeps the underlying architecture intact. Production volumes are expected to fall well short of TPU levels, and Google reportedly views the project partly as a trial run. But even as an exploratory effort, the ambition is clear.

Why Google's AI Compute Crunch Matters for GOOGL Stock

The reporting revealed something investors should pay close attention to. Google is facing an internal AI computing capacity crunch severe enough that it has fueled internal discussions and even tensions and prompted Google Cloud to decline deals with outside customers. That's not a typo. The constraint on Google's cloud growth right now is not demand. It is supply. That is consistent with everything management has said this year, from the $80 billion capital raise to fund infrastructure to the $180 billion to $190 billion capital expenditure (CapEx) guidance. A chip that serves up to 10 times more tokens per watt is a direct attack on that bottleneck, and on the single largest cost line in the AI era: inference.

How Google's TPU Strategy Strengthens Alphabet's AI Advantage

Serving efficiency is where the AI race is heading. Google already processes over 16 billion tokens per minute through its models, and every improvement in tokens-per-watt flows straight into Google Cloud's margins and CapEx efficiency. The TPU franchise has also quietly become a strategic asset in its own right, with Meta (NASDAQ: META) signing a multi-year agreement for TPU access, Anthropic pledging to draw on multiple gigawatts of TPU capacity, and all 17 U.S. Department of Energy national laboratories among its users. Frozen v2 extends that silicon advantage another generation forward.

The Intel (NASDAQ: INTC) thread makes the story bigger. In early June, it was reported that Google will rely on Intel to manufacture more than 3 million TPUs in 2028, reportedly the largest order Intel's foundry business has ever received. Then on July 16, the two companies announced an expanded partnership that deploys Gemini Enterprise across Intel's workforce and pushes agentic AI tools directly into Intel's chip design process, building on an April agreement that put Intel's Xeon 6 processors into Google's cloud. Google's custom silicon ambitions and Intel's manufacturing comeback are becoming increasingly intertwined, with American-made AI chips as the common thread.

Alphabet Earnings Put Google Cloud Growth and AI Spending in Focus

All of this lands two days before the next major catalyst. Alphabet reports Q2 2026 results on Wednesday, after the close, with analysts expecting earnings per share (EPS) of approximately $2.86 on revenue of roughly $116.5 billion, up close to 21% year over year. Google Cloud is the number to watch after growing 63% last quarter with a $462 billion backlog. And in light of Monday's report, any management commentary on capacity constraints deserves extra attention. A company turning away cloud customers for lack of compute is a company with more demand than it can monetize, which is a high-quality problem, but one investors will want to hear a plan for.

The consensus among 55 analysts is Moderate Buy with a price target of $414.11, implying double-digit upside. Monday's chip report was a reminder of why the bull case continues to thrive: Alphabet remains the only company designing frontier models, custom silicon, and hyperscale infrastructure under one roof. And Wednesday's report and commentary will show whether the numbers keep backing it up.

Where Should You Invest $1,000 Right Now?

Before you make your next trade, you'll want to hear this.

MarketBeat keeps track of Wall Street's top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis.

Our team has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on... and none of the big name stocks were on the list.

They believe these five stocks are the five best companies for investors to buy now...

See The Five Stocks Here

Featured Articles

• The Top Five Stocks to Own in July
• Alphabet Is Planning a New AI Chip. Here's Why It Matters Ahead of Earnings
• Microsoft's AI Bet Could Rewrite AMD's Growth Story
• Travelers Stock Surges 10% as Earnings Beat Reveals Underwriting Discipline
• Buy, Hold, or Wait: 3 Small-Cap Stocks Telling Different Stories
• IBM Just Had Its Worst Day Ever-What Earnings Must Prove