Intel Corporation (INTC) and Google Cloud announced an expanded partnership on Thursday 16th, deploying Gemini Enterprise across Intel’s workforce and pushing agentic AI tools into its chip design process, according to an Intel Newsroom statement.
The announcement should have read as an unambiguous win for a company trying to prove its AI credentials. Instead, Intel shares closed down 5.84% that day, sliding to $96.98 from the prior session’s $102.99, according to data from TheStreet.
The sell-off had a different author entirely.
Intel’s sell-off traces back to Taiwan, not Mountain View
Taiwan Semiconductor Manufacturing reported record second-quarter results the same Thursday, posting a 67.7% gross margin and raising its 2026 capital spending target to as much as $64 billion, up from a prior ceiling of $56 billion.
Investors read the higher capex as a sign that even the industry’s most profitable chipmaker will need years to turn AI demand into free cash flow.
Semiconductor stocks fell broadly on the news, and Intel, still trying to prove its own foundry can turn a profit, absorbed more damage than most.
Related: Does Intel pay dividends? History & future prospects explained
That single earnings report explains more of Thursday’s move than any AI partnership could. Intel had already fallen from a 52-week high of $142.35 in late June to roughly $103 by mid-July, a decline tied to reports that its 18A manufacturing process may not reach profitable yields until 2027, according to The Motley Fool.
The same week, AMD reported first-quarter data center revenue of $5.8 billion, edging past Intel’s $5.1 billion in that segment for the first time, the outlet reported. Both events mattered more to Intel’s stock than a cloud computing deal.
JHVEPhoto / Getty Images
The AI deal reaches deep into chip design
Under the expanded agreement, Intel will use Gemini Enterprise as a central platform for employees to build and deploy AI agents across engineering, supply chain and corporate operations, the Intel Newsroom statement said.
Google Cloud will also expand the computing capacity available to Intel’s chip designers, adding custom agentic workflows meant to shorten the semiconductor development lifecycle.
Google Cloud CEO Thomas Kurian said the goal is to accelerate Intel’s enterprise-wide digital transformation using Gemini Enterprise and Google Cloud.
Alphabet CEO Sundar Pichai publicly welcomed the news, framing Intel’s adoption of Gemini as a marquee example of the platform reaching into next-generation semiconductor work, Benzinga reported.
The arrangement carries a wrinkle other coverage has largely skipped. Google designs its own AI accelerators, called Tensor Processing Units, for use inside its own data centers.
Related: Morgan Stanley: Broadcom bears are wrong about Google TPU
Handing agentic chip design workflows to Google Cloud means Intel’s engineers are now building silicon partly on infrastructure controlled by a company that competes with Intel in custom AI silicon, even as it sells Intel enterprise software everywhere else.
This dynamic highlights a strange new reality in the semiconductor industry. To move fast enough to survive, companies are forced into complex relationships where they must share critical development environments with their direct rivals.
Intel is essentially betting that the speed gains from using Google’s AI will outweigh any long term strategic risks of relying on a competitor’s infrastructure.
Google shares, meanwhile, ticked up about 0.1% the same day. That muted move underscores how differently the market treats AI infrastructure providers versus AI infrastructure buyers.
Google captures steady enterprise software revenue from deals like this one regardless of how Intel’s hardware business performs.
More Intel:
- Top-rated analyst sets a jaw-dropping Intel stock price target
- Intel’s stock split history (& prospects) explained
- Intel CEO gives investors a reality check
Google needs Intel as much as Intel needs Google
This is not a one-way arrangement. Google Cloud committed to deploying Intel’s latest Xeon 6 processors across its own cloud infrastructure under an April 2026 agreement that laid the groundwork for Thursday’s expansion.
Every enterprise workload Google shifts onto Xeon hardware strengthens Intel’s case that its chips still belong inside the world’s largest cloud providers, even as Nvidia and custom silicon dominate the AI training conversation.
Enterprise AI adoption is now a side story to manufacturing risk
Wall Street increasingly treats chipmakers as two separate stories: the AI partnerships they sign, and the manufacturing execution that determines whether they can profit from them.
Intel’s Thursday captured that gap perfectly. A partnership touching nearly every part of its business barely registered against a single capital spending number out of Taiwan.
While everyday investors are reacting sharply to immediate manufacturing risks, institutional heavyweights seem willing to play a much longer game.
The U.S. government took a roughly 10% equity stake in the company last August, and Nvidia invested $5 billion in December even though the two chipmakers compete for the same AI infrastructure dollars, according to The Motley Fool.
Google’s expanded AI partnership fits that same pattern: strategic partners keep betting on Intel’s turnaround even as the stock keeps testing investors’ patience.
Intel reports second-quarter earnings on July 23. That report, not this week’s Google Cloud news, will show whether Intel’s foundry turnaround can keep pace with the AI story it keeps telling investors.
Specifically, Wall Street will be looking for concrete updates on the 18A manufacturing timeline and any signs of improvement in the company’s foundry margins.
If management cannot prove they are controlling the physical costs of building these chips, no amount of AI software integration will be enough to calm nervous shareholders.
Related: Intel surge hints far beyond Apple news
