Intel (INTC) surged roughly 11% on Monday after reports that Google (Alphabet) selected Intel Foundry as a secondary manufacturer for more than 3 million next-generation TPU chips, targeting 2028 production. The order is the largest manufacturing contract Intel Foundry has ever secured and its first major production agreement with a top-tier cloud provider (Reuters).

The news provides powerful external validation for CEO Lip-Bu Tan's foundry strategy, which he outlined in May when he disclosed a pipeline of 20+ potential customer deals worth over $15 billion in lifetime value (CNBC). Intel now trades around $110 — more than double its 200-day moving average of $51.71 — completing a staggering ~199% year-to-date return.

Google's decision was not impulsive. Alphabet has been testing Intel's advanced packaging technologies — particularly EMIB (Embedded Multi-die Interconnect Bridge) and Foveros 3D stacking — for months. According to earlier reports from WCCFTech, Google was already using Intel for advanced packaging of test TPU designs since April 2026. The production order expands this to mass-manufacturing scale involving millions of units (WCCFTech).

Why It Matters

The global semiconductor industry is undergoing a structural shift that could reshape the competitive landscape for the next decade. TSMC currently controls roughly 90% of advanced chip manufacturing — a concentration risk that creates enormous geopolitical exposure, especially given U.S.-China tensions and the existential threat to Taiwan.

Major customers like Google, Apple, Amazon, Microsoft, and Nvidia are actively seeking alternatives. Intel's 18A process offers a U.S.-based alternative, backed by roughly $7.9 billion in CHIPS Act grants and explicit government pressure to diversify supply chains. The Google order demonstrates that this strategy is beginning to produce tangible results.

Beyond Google, reports indicate that Nvidia is evaluating Intel's 18A process as a potential backup manufacturer for its next-generation Feynman GPU architecture (Bloomberg). The U.S. government has also explicitly encouraged Apple to diversify away from TSMC. These converging signals point to a broader "TSMC-plus-one" trend — leading chip designers wanting a reliable American backup supplier.

The Bull Case

Foundry validation at scale. Google is the most sophisticated AI chip customer after Nvidia. Choosing Intel as a secondary supplier for 3 million units is the strongest signal yet that Intel's 18A process is competitive with TSMC at the leading edge. The order also leverages Intel's EMIB advanced packaging and Foveros 3D stacking — distinct competitive advantages that hyperscalers increasingly need for complex multi-die AI designs.

A rapidly deepening customer pipeline. Beyond Google, Intel has multiple marquee names in various stages of engagement. A preliminary foundry agreement with Apple would mark one of the most significant validation points possible — Apple is TSMC's largest customer and the industry's most demanding design house (WSJ). Intel joined the Terafab project with Tesla, SpaceX, and xAI as a manufacturing partner for AI inference chips (AI5/AI6) and radiation-hardened space processors on 18A and 14A nodes (Wikipedia). A partnership with Foxconn aims to accelerate AI infrastructure deployment (Barron's). And Nvidia evaluating Intel 18A for its next-gen Feynman architecture adds a second powerful catalyst.

U.S. policy tailwind is real and durable. The CHIPS Act, combined with bipartisan pressure to onshore semiconductor manufacturing, creates a uniquely favorable environment for Intel. Unlike foreign competitors, Intel has access to government grants, preferred contracting from defense and intelligence agencies, and explicit encouragement from Washington for American tech companies to use Intel as a foundry partner.

Yield improvement trajectory is encouraging. Intel reports 18A yields are improving roughly 7% per month. If the trend holds, commercially competitive yield levels (65-75%) are within reach in the coming months — a critical milestone for foundry profitability.

Positive free cash flow on the horizon. Intel guided for positive adjusted free cash flow in 2026, the first time in years. Revenue is growing 7.2% YoY to $53.8B, and the current ratio of 2.3 indicates healthy liquidity (Trefis).

The Bear Case

Google revenue is a 2028 story — the stock has already priced it in. At $110 with a forward P/E of 71.6x on still-negative profit margins (-5.9%) and negative ROE (-2.9%), Intel trades 24% above the mean analyst consensus target of $89 (range: $20.40 – $150 across 41 analysts). The rally from $19 to $110 has been a multiple-expansion story, not an earnings story. Revenue per share is $11.40 — at $110, the stock trades at roughly 9.7x revenue (MarketBeat).

Intel is Google's secondary supplier, not primary. Google continues to manufacture the bulk of its TPUs at TSMC. According to Semicone, Intel will handle roughly half of Google's total 2028 production volume, estimated at over 6 million units. This is about TSMC capacity constraints and supply-chain diversification, not about Intel being technologically superior. The "TSMC-killer" narrative is overblown (Semicone).

Execution risk on 18A is existential. Intel has never operated as a merchant foundry at industrial scale. It has historically manufactured primarily for its own designs — a fundamentally different business from being a reliable contract manufacturer for external customers demanding absolute precision and on-time delivery. TSMC's CEO has publicly warned that "no shortcuts exist" in building and ramping advanced fabs — TSMC rests on 40 years of merchant foundry experience (Tom's Hardware). If 18A yields disappoint, the entire foundry thesis unravels.

TSMC is not standing still. TSMC N2 is expected to be competitive with Intel 18A, and TSMC has the scale, track record, and customer relationships that Intel is just beginning to build. TSMC is rapidly adding capacity, which could reduce the urgency for customers to diversify.

The rally may have overshot fundamentals. INTC is up 199% YTD, trading at 2.1x its 200-day moving average. Rosenblatt maintains a Sell rating with a $30 target; JP Morgan is Underweight at $30. The analyst consensus is Hold with a mean target 24% below the current price. The 11% single-day pop on a deal that generates revenue in 2028 raises concerns about "buy the rumor, sell the news" dynamics and retail euphoria.

High debt load. Intel's debt-to-equity ratio stands at 36x — an extremely high figure that limits financial flexibility in case of a downturn or delays in the foundry ramp.

What to Watch

Q2 earnings (July 23, 2026). The upcoming earnings report will be a critical test. The market will scrutinize three metrics above all: 18A yield progress, foundry customer commitment updates, and free cash flow trajectory. A beat-and-raise could justify the elevated valuation; a miss could trigger a sharp correction from these levels.

Apple deal finalization. The preliminary chip-making agreement with Apple is arguably Intel's next biggest opportunity. If it converts to a binding contract with defined volume commitments, it would add another powerful validation point — and prove Google is not a one-off.

Additional hyperscaler wins. Amazon (AWS), Microsoft (Azure), and Meta are all exploring supply-chain diversification. Any one of them signing with Intel Foundry would be a major catalyst for the "TSMC-plus-one" narrative.

Nvidia Feynman decision. If Nvidia formalizes its evaluation into actual purchase commitments for the Feynman architecture, it would add credibility beyond the Google deal and signal a genuine industry shift.

14A node decision. Intel's late-2026 decision on whether to proceed with full-scale 14A capacity investment (targeting 2028 volume production) will signal management's confidence in sustained demand. A go-ahead would be a major vote of confidence from within.

The Bottom Line

The Google deal is a genuine milestone for Intel. It validates Lip-Bu Tan's strategy and positions Intel as a legitimate player in advanced foundry services after years of skepticism. But at $110, the stock already reflects much of that optimism — and then some.

Intel in 2026 is a multiple-expansion story, not an earnings story. The stock has risen from $19 to $110 primarily on multiple expansion, not on improvement in underlying profitability. If Intel executes flawlessly on 18A and signs additional anchor customers, the stock can go higher. But at 71.6x forward earnings on still-negative margins, any execution hiccup risks a painful re-rating.

Investors should separate the strategic validation — which is real and meaningful — from the valuation, which already prices in years of successful execution. This is a stock for patient investors with multi-year horizons, not for those chasing quick gains. The July earnings report will be the first real test of whether Intel is genuinely turning the corner, or whether this is another false dawn.