By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
The panel is divided on Intel's EMIB/EMIB-T packaging and Google's TPU order, with concerns about long-term execution, yield issues, and high valuation outweighing the validation of Intel's packaging technology.
Risk: Long-term execution risk, including customer concentration, unit economics, and yield compression.
Opportunity: Potential geopolitical hedge as an alternative to TSMC's Taiwan-centric supply chain.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
After years of losing the plot, Intel (INTC) seems to be gaining back its mojo. While skeptics may dismiss it as a government-led bailout of the beleaguered chip giant, Intel is slowly stacking up the “W's”. This time it is from Alphabet's Google (GOOGL) (GOOG).
According to a report on The Information, Intel has secured an order to build three million TPU chips for Google. To be manufactured in 2028, Intel's advanced packaging capabilities worked in favor of the chip major for this order.
Shares reacted positively to the news, rallying by more than 11% in Monday's trading session. Notably, INTC stock is already up 191.72% year-to-date (YTD) and holds a market cap of $554 trillion dollars, a benchmark unthinkable for the company just about five years ago.
Diversification Not The Sole Reason
Several market experts are citing diversification from the world's largest manufacturer of chips, TSMC (TSM), as the reason behind this move from Google. There is definitely substance to this theory, but it doesn't represent the entire picture.
The packaging angle is where the real logic sits, and it deserves more attention than it typically receives. Intel's core technology is EMIB, its Embedded Multi-die Interconnect Bridge, which is a silicon bridge embedded directly in a substrate that provides high-density die-to-die connectivity without the cost and area overhead of a full-size silicon interposer. TSMC's equivalent is CoWoS, and the problem is simple. TSMC's CoWoS capacity remains tightly constrained, and Intel's EMIB yields have reportedly reached approximately 90%.
Moreover, what makes EMIB genuinely attractive to Google is its HBM integration story. Intel's EMIB-T variant, which adds through-silicon vias to the bridge, supports scaling beyond 8 times reticle size in a 120 by 120 millimeter package configuration, enabling integration of up to 12 HBM stacks and four high-density chiplets, which is exactly the kind of memory bandwidth architecture that Google's TPU designs demand. Google's TPU v8e, scheduled for the second half of 2027, is expected to leverage Intel's EMIB technology, meaning this manufacturing order can be one of many in the future.
Beyond packaging, other technical reasons are worth considering. Google has been increasingly aggressive about owning its AI infrastructure stack, and a multi-generation dependency on a single foundry for its most critical custom silicon carries real supply chain risk at the scale Google operates. Intel CFO David Zinsner noted on the Q1 2026 earnings call that the advanced packaging business revenue potential is now expected to exceed $1 billion, with EMIB and EMIB-T customers expected as early as the second half of 2026, bringing several billion dollars in revenue, which confirms that Google's engagement is part of a broader hyperscaler validation cycle rather than a standalone deal.
There is also a cost dimension. EMIB avoids the full interposer required by CoWoS, which structurally reduces packaging costs per unit at the volumes Google is ordering.
Numbers On The Right Path
Intel delivered a strong start to 2026 with a clear double beat on both revenue and profit figures during the first quarter.
Total revenue rose 7% year-over-year (YOY) to $13.6 billion. The main drivers were the Data Center and Artificial Intelligence segments, together with the Foundry business. The Data Center and Artificial Intelligence unit grew 22% to $5.1 billion, and the Foundry segment advanced a solid 16% to $5.4 billion. Meanwhile, the Client Computing Group, which remains Intel’s biggest revenue contributor, posted only modest growth of 1% to $7.7 billion, held back by supply issues and a saturated personal computer market.
On the earnings side, non-GAAP earnings per share more than doubled to $0.29, beating consensus estimates by a healthy margin. This result extended the company’s run of consecutive earnings beats to three quarters.
Management guided for second quarter revenue in the range of $13.8 billion to $14.8 billion, with non-GAAP earnings per share expected at $0.20. Wall Street currently anticipates $14.39 billion in revenue and $0.21 per share.
Also, cash flow strengthened, as net cash from operating activities increased to $1.1 billion from $813 million in the year ago period. Intel closed the quarter with a robust cash balance of $17.7 billion, comfortably exceeding its short term debt of approximately $2 billion.
Valuation continues to represent a notable challenge for the INTC investment story. The shares trade at a premium to both industry peers and the company’s own historical norms. The forward price-to-earnings of 101.11 times, forward price-to-sales of 9.43 times, and price-to-cost flow of 44.22 times all stand well above sector medians. As a result, the stock depends heavily on consistent operational execution to validate the higher valuation the market has placed on it.
Analyst Opinion
Overall, analysts have deemed a rating of “Hold” for INTC stock, with the mean target price already surpassed. However, the Street-high target price of $150 indicates potential upside of 37.35% from current levels. Out of 44 analysts covering INTC stock, nine have a “Strong Buy” rating, one has a “Moderate Buy” rating, 31 analysts have a “Hold” rating, one has a “Moderate Sell,” and two have a “Strong Sell” rating.
On the date of publication, Pathikrit Bose did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. This article was originally published on Barchart.com
Four leading AI models discuss this article
"Intel's advanced packaging (EMIB/EMIB-T) could become a multi-year, material revenue engine if the Google TPU deal scales, potentially re-rating INTC on AI-packaging leverage."
Intel’s embrace of Google’s AI ambitions via EMIB/EMIB-T packaging signals real progress beyond headline diversification. The real lever isn’t the three million TPUs on a 2028 ramp, but whether Intel can scale advanced packaging, deliver acceptable yields, and meaningfully lower Google’s single-foundry risk. However, the article glosses over critical caveats: a) 2028 production is a long horizon with potential delays; b) Intel’s historical packaging cost/yield challenges and overall AI infrastructure demand could outstrip supply or wallet share; c) the claimed market-cap and some numbers in the write-up look inflated, which dampens credibility. The bull case rests on durable packaging growth and true 1B+ revenue from EMIB/EMIB-T, not a one-off contract.
The strongest counter is that this may be a small, long-horizon pilot rather than a durable catalyst. Intel’s history of cost overruns and yield challenges in advanced packaging makes a 2028 ramp precarious, and the Google dependence risk may linger.
"The 101x forward P/E ratio is unsustainable for a legacy chipmaker undergoing a high-capex transition, regardless of individual hyperscaler wins."
The market is overreacting to a 2028 manufacturing order. While Intel’s EMIB and EMIB-T packaging technology is genuinely superior for high-bandwidth memory integration, the stock’s current valuation—trading at a staggering 101x forward P/E—is detached from the reality of its foundry execution risks. Intel is currently priced for perfection, yet its Client Computing Group remains stagnant. The Google deal is a validation of Intel's packaging R&D, but it doesn't solve the structural margin compression Intel faces as it pivots from a high-margin IDM model to a lower-margin foundry business. Investors are ignoring the massive capital expenditure required to scale this capacity, which will likely weigh on free cash flow for years.
If Intel successfully scales EMIB-T as the industry standard for hyperscaler AI chips, they could capture enough market share from TSMC to justify a premium valuation, effectively becoming the 'Intel Inside' of the AI era.
"Intel's packaging breakthrough is real, but a 101x forward P/E leaves no margin of error for the execution risk that has defined the company's last five years."
The TPU deal is real validation of Intel's packaging tech, but the article buries the valuation problem. INTC trades at 101x forward P/E—absurdly expensive for a company that just posted 7% YoY revenue growth. The Foundry segment is 40% of revenue but still losing money (article omits this). Google's 3M TPU order, even at scale, won't move the needle enough to justify current multiples. The 191% YTD rally has already priced in a turnaround; execution risk is now asymmetric to the downside. Three earnings beats don't erase years of process node delays.
If Intel's EMIB-T truly solves Google's HBM bandwidth bottleneck and hyperscalers validate it across multiple generations, this becomes a multi-billion-dollar recurring revenue stream that could justify premium multiples—and the market is still early in pricing that optionality.
"A 2028 manufacturing order and stretched valuation leave Intel's re-rating dependent on near-term foundry wins the article does not address."
The Google TPU order validates Intel's EMIB-T packaging edge over TSMC's constrained CoWoS capacity and supports multi-year hyperscaler revenue, yet the 2028 delivery timeline means no material 2026-2027 impact. Intel's Q1 foundry growth and $1B+ packaging guidance are real, but the 101x forward P/E already prices in flawless execution across a loss-making foundry business still reliant on external customers. Historical yield and process misses suggest this single deal does not resolve structural foundry competitiveness versus TSMC.
The order could trigger follow-on TPU v8e and v9 volumes starting 2027, accelerating the several-billion-dollar packaging pipeline Zinsner flagged and forcing a re-rating before the high multiple compresses.
"The real catalyst risk is durability of multi-customer, high-margin EMIB-T revenue; a stall in adoption or reliance on Google alone could keep the stock vulnerable to multiple compression."
Grok's push that 2027 TPU volumes could trigger a re-rate assumes a clean, multi-year ramp in a single packaging platform. The bigger, unaddressed risk is customer concentration and unit economics: even if packaging wins, a few large orders don't guarantee durable margins as capex and yields compress returns. If Google becomes the primary, or if EMIB-T adoption stalls, the 'packaging pipeline' may stay small and the multiple remains vulnerable.
"Intel’s valuation premium is driven by geopolitical supply chain diversification rather than just foundry operational efficiency."
Gemini and Claude highlight the 101x forward P/E, but you are all missing the geopolitical hedge. Google is not just buying packaging; they are buying an alternative to TSMC’s Taiwan-centric supply chain. If the market prices in 'sovereign silicon' risk, that 101x multiple isn't just about earnings—it’s a premium for supply chain insurance. Intel doesn't need to beat TSMC on yield; they just need to be the only viable non-Taiwanese option for hyperscalers.
"Supply-chain diversification is a real Intel tailwind, but it doesn't override the valuation math if execution risk remains high."
Gemini's geopolitical hedge argument is seductive but incomplete. Yes, supply-chain insurance commands a premium—but only if Intel can actually *deliver* at scale without yield collapses. TSMC's Taiwan risk is real, yet Intel's foundry track record suggests they may be the riskier counterparty operationally. The 'sovereign silicon' premium evaporates if Google ends up qualifying SAMSUNG or GlobalFoundries as backup. Geopolitics alone doesn't justify 101x forward P/E without execution credibility.
"Geopolitical diversification alone cannot sustain Intel's premium valuation without matching TSMC on execution metrics."
Gemini’s geopolitical hedge claim assumes hyperscalers will pay up for Intel solely as a Taiwan alternative, yet Google still qualifies multiple foundries and will demand competitive yields and costs. Without closing that gap, the sovereign-silicon premium evaporates and the 101x forward P/E reverts to reflecting chronic foundry losses and capex intensity rather than durable packaging revenue.
The panel is divided on Intel's EMIB/EMIB-T packaging and Google's TPU order, with concerns about long-term execution, yield issues, and high valuation outweighing the validation of Intel's packaging technology.
Potential geopolitical hedge as an alternative to TSMC's Taiwan-centric supply chain.
Long-term execution risk, including customer concentration, unit economics, and yield compression.