Intel Rebounds 4.9% to $116 as the Products Engine Carries the 18A Foundry Dream — $125.70 Is the Breakout to the Record

Intel is rebounding hard on Thursday, up roughly 4.9% to around $116 with an intraday high of $116.77, as dip-buyers step back into the chip sector after a punishing selloff. The bounce follows a brutal stretch that erased 21% from the stock in seven days, dragging it from near its $142 record down to $110 on a wave of semiconductor profit-taking and fresh doubts about Intel's foundry timeline. The recovery puts the stock back within striking distance of its 52-week high of $142.35, and it caps one of the most extraordinary re-ratings in the market: Intel has climbed more than five-fold over the past year from around $19, adding close to $500 billion in market value for a company that recently posted a $16.6 billion GAAP loss and cut 15,000 jobs.

The July 9 rebound has clear drivers. A broad chip-sector recovery took hold as SK Hynix and Kioxia jumped 8% overnight and dip-buyers rotated back into semiconductors ahead of SK Hynix's $28 billion Nasdaq listing, whose ADRs begin trading July 10 more than seven times oversubscribed, a signal of institutional confidence in the AI semiconductor supply chain. HSBC's Wall Street-high $200 price target, set days ago with a maintained Buy rating on undervalued data-center CPU and foundry businesses, continues to anchor bullish sentiment. Intel's own confirmed price increases on consumer and Xeon server processors reinforce the revenue outlook ahead of its Q2 earnings on July 23.

That sets up the thesis: Intel is a two-engine stock where the market keeps confusing which engine is actually running. The real, profitable Products business, data center up 22%, server CPU pricing power, genuine earnings, is doing the work that justifies the price, while the speculative Foundry turnaround built on the 18A process is doing the dreaming. The 21% selloff was the market violently repricing the gap between the two after reports that 18A won't hit profitable yields until 2027, and the July 9 bounce is dip-buyers betting the Products engine carries the valuation even if the Foundry engine is years from profit. The bracket runs from $125.70 above, toward the $141 record, to $108.60 and $100 below. The whole story resets on July 23 earnings, where external foundry revenue is the number that matters. This is a genuine recovery layered over a speculative dream, and the two keep getting conflated.

Two Engines: One Running, One Dreaming

The key to understanding Intel is recognizing that its valuation rests on two separate stories that the market constantly blends together, and separating them is the whole forecast. The first engine is Intel Products, the actual chip business selling CPUs today, which is genuinely recovering and generating real revenue and earnings. The second engine is Intel Foundry, the contract-manufacturing turnaround built on the 18A process node, which is a speculative bet on a future that has not yet arrived. Intel's five-fold re-rating over the past year came from a combination of both, and the confusion between them is why the stock is so volatile.

The Products engine is running. Data Center and AI revenue jumped 22% year-over-year to $5.05 billion, the client CPU business runs around $31 billion annually, server CPU revenue grew 20% to 25% driven by pricing, and Intel confirmed fresh price hikes on consumer and Xeon processors that show real demand exceeding supply. This is a profitable, cash-generating business with genuine AI-driven tailwinds, and it is the part of Intel that justifies a substantial portion of the current price. When analysts talk about Intel's earnings today, they are talking about Products.

The Foundry engine is dreaming. The 18A process is in volume production, but external foundry revenue was just $174 million in Q1 against a $2.4 billion operating loss, which means the foundry is still overwhelmingly serving Intel itself rather than winning outside customers. The bull case for foundry, that Intel becomes a credible TSMC alternative winning Apple, Google, and Tesla, is a multi-year bet that depends on execution not yet proven. For the forecast, the two-engine framing is essential: the Products engine is real and carries the near-term valuation, while the Foundry engine is speculative and carries the long-term upside and the biggest risk. The 21% selloff happened when the market got scared about the Foundry dream, 18A profitability delayed to 2027, and the July 9 rebound happened when it refocused on the Products reality, server CPU demand and pricing. The stock trades on whichever engine the market is watching, and the July 23 earnings will force a reckoning on both: Products results plus the critical external foundry revenue number. Intel is one running engine plus one dreaming engine, and the price depends on believing both.

The July 9 Rebound Rests on Three Pillars

The specific catalysts behind Thursday's 4.9% bounce reveal what the market is focused on, and the first pillar is the broad semiconductor recovery. After two brutal sessions of profit-taking, dip-buyers returned to the AI hardware trade, with Kioxia rising 8.3% in Japan and Samsung and SK Hynix gaining as investors rotated back into memory names. The rebound had traction because analysts did not treat the pullback as a break in the AI cycle, viewing it instead as valuation-driven profit-taking in an intact structural uptrend. When the whole chip complex bounces, Intel bounces with it, and the SK Hynix listing provided the catalyst.

The second pillar is the SK Hynix listing itself. SK Hynix's planned $28 billion Nasdaq sale drew demand more than seven times oversubscribed, and its ADRs begin trading July 10, a signal that institutional confidence in the AI memory and semiconductor supply chain remains strong despite the week's turbulence. A heavily oversubscribed mega-listing tells the market that big money still wants semiconductor exposure, which lifts sentiment across the sector and supports Intel as part of the AI supply chain. The listing is a vote of confidence in the entire chip trade.

The third pillar is Intel-specific: HSBC's $200 target and the price hikes. HSBC analyst Frank Lee doubled his Intel target to $200 from $100, calling the foundry push too good to ignore and citing undervalued data-center CPU and foundry businesses, and that Wall Street-high target continues to anchor bullish sentiment. Intel's confirmed price increases on consumer and Xeon processors, justified by supply shortages and AI-driven demand exceeding supply, reinforce the revenue outlook ahead of July 23. For the forecast, the July 9 rebound rests on sector recovery, the SK Hynix confidence signal, and Intel-specific bullish catalysts in HSBC's target and the price hikes. The bounce is the market re-embracing the Products story and the foundry optionality after the valuation scare. But the rebound only erased part of the 21% loss, which shows the doubts linger, and the durability of the recovery depends on the July 23 earnings validating the bull case. The three pillars lifted Intel off its lows, but the reset is still ahead.

The Products Engine Is Genuinely Recovering

The foundation under Intel's valuation is the Products business, and it is delivering real results driven by AI demand for CPUs. First-quarter 2026 revenue rose 7% year-over-year to $13.6 billion with non-GAAP EPS of $0.29, and the standout was Data Center and AI, where revenue jumped 22% year-over-year to $5.05 billion, beating expectations of $4.41 billion. That data-center business now runs at roughly $20 billion annualized and is growing, driven by the shift toward agentic AI that CEO Lip-Bu Tan says is increasing demand for advanced CPUs and wafer-packaging technologies. The AI-is-good-for-CPUs-again narrative is showing up in the actual numbers.

The server CPU story is the core of the recovery. Server CPU revenue grew in the 20% to 25% range in Q1, mostly through higher average selling prices, and Intel's Xeon 6 processors have been selected for high-profile AI server architectures, with a growing partnership base among major hyperscalers. As AI data centers expand, they need CPUs to orchestrate the workloads alongside the GPUs, and Intel's Xeon franchise is capturing that demand. CFO David Zinsner described AI as creating explosive growth in the CPU market, a striking characterization for a business many had written off.

The forward guidance reinforces the momentum. Intel raised its outlook in April, guiding second-quarter revenue to $13.8 billion to $14.8 billion, ahead of the $13.07 billion analysts expected, and adjusted profit of $0.20 per share, more than double the $0.09 estimate. Tan told analysts that the strong outlook reflects what the company hears from customers, not wishful thinking. For the forecast, the Products engine is the real, profitable business that justifies a substantial part of Intel's valuation. Data center up 22%, server CPU up 20% to 25%, Xeon 6 in AI servers, and guidance well above expectations show a genuine recovery driven by AI CPU demand. This is the engine doing the work, and it is why dip-buyers returned on July 9. The bull case for the near term rests here, on the Products business converting AI demand into revenue and earnings, and the July 23 report will show whether the recovery is accelerating. The Products engine is running, and it is carrying the stock.

Pricing Power Is Doing the Heavy Lifting

A crucial nuance in Intel's recovery is that much of the server CPU growth is coming from price increases rather than volume, and that pricing power is a genuine strength with a catch. Intel confirmed selective price hikes on specific consumer and high-end Xeon server processors on July 6, justified by supply shortages and AI-driven demand exceeding supply, and the server CPU revenue growth of 20% to 25% in Q1 came mostly through higher average selling prices. That Intel can raise prices and still sell chips reflects real demand strength and tight supply, which is a positive signal for margins and revenue.

The pricing power stems from a supply-demand imbalance in server CPUs. AI data centers need CPUs faster than the industry can supply them, which lets Intel command higher prices, and the price hikes flow almost entirely to the bottom line since they require no additional volume. This is the kind of pricing environment that lifts margins quickly, and it is part of why Intel's Q2 EPS guidance of $0.20 more than doubled the prior estimate. Pricing power is the cleanest path to margin recovery for a company that has struggled with profitability.

The catch is that price hikes can erode competitiveness. Raising prices helps margins in the near term, but it may make Intel's chips less competitive against alternatives like AMD's, which could accelerate the share loss Intel is already suffering in the data center. If customers can get comparable performance from AMD at a lower price, Intel's price increases could push them toward the competitor, trading near-term margin for long-term share. For the forecast, the pricing power is a double-edged strength: it drives the near-term revenue and margin recovery that supports the stock, but it carries the risk of accelerating competitive share loss if AMD and Arm undercut Intel on price. The fact that so much of the growth is ASP-driven rather than volume-driven is a yellow flag, because volume growth is more durable than price growth. Intel is monetizing a tight supply environment effectively, but the pricing strategy works only as long as demand exceeds supply and customers accept the higher prices. The July 23 earnings will reveal whether the pricing power is holding or whether volume is slipping to competitors, a key detail for the durability of the Products recovery.

The 18A Foundry Dream Is Real but Early

The speculative engine driving Intel's long-term upside is the foundry turnaround built on the 18A process, and it has made genuine progress that justifies some of the optimism. Intel's 18A process is no longer just a promise, it is in volume production: in January 2026, Fab 52 in Chandler, Arizona started making 18A chips at high volume, with output around 40,000 wafers per month and reported yields of 65% to 75%, good enough for commercial production on a new advanced node. The first 18A products, Panther Lake laptop chips and Clearwater Forest server CPUs, shipped in late January, and the next version, 18A-P, has entered risk production with 9% higher performance or 18% lower power.

The strategic significance is that 18A could restore Intel's manufacturing leadership. For years Intel fell behind TSMC in process technology, which was the root of its decline, and a competitive 18A node would let Intel both make better chips for itself and win external foundry customers, the two-sided payoff that underpins the bull case. Intel has $15 billion in signed lifetime foundry commitments, and reports of potential partnerships with Apple, Google, Tesla, and SpaceX have improved confidence in the manufacturing roadmap. The appointment of former SK Hynix CEO Seok-Hee Lee to lead advanced packaging adds credibility to the foundry's technical leadership.

The critical caveat is that 18A profitability is delayed. Reports indicate that the 18A and 18A-P processes may not achieve profitable commercial yields until late 2026 or 2027, which is precisely what triggered the recent selloff, because it pushes out the high-margin foundry payoff that the valuation assumes. Yields on the Panther Lake compute tile are reported in the 50% to 60% range by some industry estimates, below the threshold for profitable scale manufacturing, with improvement expected as production matures. For the forecast, the 18A foundry dream is real, it is in production with decent yields, has a product roadmap, and has signed commitments, but it is early and not yet profitable. The dream is the source of Intel's long-term upside and the reason for targets like HSBC's $200, but the delayed profitability is the source of the near-term risk and the recent selloff. The foundry engine is dreaming of a profitable future that the market keeps trying to price in and then getting scared about. July 23 is where the dream meets the numbers.

The Foundry Reality Is $174 Million and Heavy Losses

Beneath the 18A optimism lies a sobering financial reality: Intel Foundry is still overwhelmingly an internal supplier bleeding money, and this is the bear case's strongest point. Intel Foundry achieved $5.4 billion in revenue in Q1 2026, but external foundry revenue, the money from outside customers that actually validates the turnaround, was just $174 million. The rest came from Intel manufacturing chips for itself, which verifies the technology but does not prove the foundry can win the external business that justifies the valuation. For all of 2025, total foundry revenue was $17.8 billion with external contributions of a mere $307 million, alongside a $10.3 billion operating loss.

The losses are the immediate problem. Intel Foundry drew approximately $2.4 billion in operating losses last quarter against only $174 million in external revenue, remaining largely unchanged from the $2.3 billion deficit the prior year, and those losses pressure Intel's overall margins amid massive ongoing capital investment. Building leading-edge fabs costs tens of billions, and until external customers commit to large production volumes, the foundry keeps consuming cash without generating the returns to justify it. The financial viability of a foundry only improves when external clients trust the process enough to commit to large volumes, and that has not happened yet.

The distinction between commitments and revenue is where the risk concentrates. Intel touts $15 billion in signed lifetime foundry commitments, but commitments are not production revenue, and the strategic uncertainty Intel created, shifting 18A from internal-only to external and back, makes customers cautious about multi-billion-dollar, multi-year manufacturing bets that require process stability. For the forecast, the foundry reality is the anchor on the bull case: $174 million in external revenue against $2.4 billion in quarterly losses shows the turnaround is far from proven, and the stock's valuation prices in a foundry success that the financials do not yet support. This gap between the foundry dream and the foundry reality is exactly what the 21% selloff repriced, and it is why the July 23 external foundry revenue number is the single most important figure in the report. If external revenue moves meaningfully above $174 million, the turnaround gains credibility; if it stays flat, the dream looks expensive. The foundry reality is heavy losses and minimal outside revenue, and it is the bear's best argument.

The 21% Selloff Was a Repricing of the Dream

The violent seven-day decline that preceded Thursday's rebound is instructive, because it shows exactly what scares Intel investors. The stock dropped 21% from around $140 to $110, and the selloff had three triggers. It started on July 1 when Bank of America warned that AI semiconductor valuations had become too high, a bubble-risk call that pulled Intel down 8.32% alongside AMD, TSMC, and Nvidia. The selling continued with a 9% drop on July 2 and 5% on July 3 as the sector-wide profit-taking gathered momentum.

The Intel-specific blow came on July 7. The stock tumbled 9.87% after reports that the 18A and 18A-P processes would probably not achieve profitable yields until late 2026 or 2027, striking directly at the foundry turnaround thesis that underpinned the valuation. Because the primary investment case rested on the foundry business rapidly securing large-scale commitments, any perceived pushback in the manufacturing timeline severely penalized the stock. The delay reframed the foundry dream as further away than the market had priced.

The competitive blow followed on July 8, when the stock fell another 3.62% on revelations that AMD had surpassed Intel in quarterly data-center revenue for the first time in modern history. That historical crossover raised concerns that Intel is losing critical ground in the lucrative enterprise server market, even as it implements price increases. For the forecast, the 21% selloff was a repricing of the gap between Intel's valuation and its execution reality: a valuation scare (BofA bubble), a foundry timeline scare (18A delay), and a competitive scare (AMD crossover) combined to knock the stock down. The selloff revealed that Intel's premium price is vulnerable to any bad news on the foundry or competitive fronts, because the valuation assumes flawless execution. The July 9 rebound erased about half the loss, showing dip-buyers still believe in the Products engine, but the speed and depth of the decline is a warning about how much optimism is priced in. The stock can fall 21% in a week on execution doubts, which is the risk embedded in the re-rating.

The AMD and Arm Share Erosion Is Structural

The most concerning fundamental development for Intel is the erosion of its dominance in server CPUs, and it is structural rather than cyclical. Intel's server CPU market share declined to 54.9% in the first quarter from 64.4% previously, while AMD expanded its share to 27.4% and Arm Holdings increased to 17.7%. That is a dramatic loss of nearly 10 points of share in the highly profitable server market, and it culminated in AMD surpassing Intel in quarterly data-center revenue for the first time in modern history, a symbolic milestone that underscores Intel's weakening competitive position.

The threat comes from two directions. AMD continues to take share with competitive EPYC processors benefiting from strong AI infrastructure demand, and its data-center crossover shows it is now outgrowing Intel in the segment that matters most. At the same time, Arm-based systems are gaining adoption among hyperscale cloud providers, reflecting a broader structural shift as hyperscalers design their own Arm-based chips to reduce reliance on x86. Arm's rise to 17.7% of server share represents a fundamental architectural challenge that Intel's x86 franchise cannot easily counter, since it involves customers moving to a different chip architecture entirely.

The significance is that Intel is defending its core profit engine against intensifying competition on two fronts. The Products recovery that supports the stock depends on the server CPU business, and if AMD and Arm keep taking share, the data-center growth that drove the 22% increase becomes harder to sustain. Intel's price increases, while helping margins, could accelerate the share loss by pushing price-sensitive customers toward AMD. For the forecast, the AMD and Arm share erosion is a structural headwind that undercuts the durability of the Products recovery. Intel's server CPU dominance, once 64.4%, is eroding toward 55% and falling, and the AMD data-center crossover is a warning that the competitive position is weakening even as AI demand lifts the whole market. The bull case assumes Intel holds enough share to grow with the market; the bear case is that AMD and Arm keep taking share until Intel's growth stalls. This is the competitive risk beneath the Products engine, and it is why the market watches Intel's share numbers as closely as its revenue. The share erosion is real, structural, and accelerating, and it caps the Products bull case.

The Apple Deal and Government Stake Are the Optionality

Two unique factors give Intel optionality that most chipmakers lack: a potential Apple manufacturing deal and a U.S. government equity stake. On the Apple front, The Wall Street Journal reported in May 2026 that Intel had a preliminary deal to make some Apple Silicon chips on 18A-P, and President Trump publicly confirmed the agreement on June 18, sending the stock up 13.9% on the day. Any Intel role in making Apple chips would be a major validation of 18A, since it would mean the most demanding customer in consumer silicon trusts Intel's process, and it would represent a step toward diversifying high-end manufacturing away from TSMC.

The reality of the Apple deal is more modest than the reaction suggested. Based on supply-chain analysis, the deal appears limited to entry-level M-series chips for the base MacBook Air and iPad Pro, with expected volume of 15 to 20 million units a year and mass production likely in late 2027, not 2026. Apple's iPhone chips and higher-end Mac chips are expected to stay with TSMC, and neither Apple nor Intel has officially confirmed the arrangement. So the Apple deal is a genuine positive and a credibility marker, but it is smaller and further out than the 13.9% stock jump implied.

The government stake adds a structural floor. In August 2025, the Trump administration converted approximately $8.9 billion in CHIPS Act grants into equity, taking a 9.9% non-voting stake in Intel, which effectively removed the tail risk that Intel foundry fails completely and the stock visits $10. That government backing as a national-security asset is part of why smart money bought Intel in the $20s and $30s, not because they believed the CPU narrative but because the catastrophic-failure scenario had been structurally removed. For the forecast, the Apple deal and government stake are the optionality that distinguishes Intel: the Apple deal is a real but modest validation of 18A that could grow, and the government stake provides a floor that reduces downside risk. Together they support the long-term bull case and explain part of the re-rating from $19. But the Apple deal is late-2027 and limited, and the government stake is a floor, not a growth driver. The optionality is real and valuable, but it is not enough to justify the current valuation on its own, which still depends on the Products engine and the foundry execution.

That's TradingNEWS