Following the market close on April 23rd, Intel surged 20%, with its stock price approaching $80, reaching its highest level since the dot-com bubble.
However, the magnitude of this gain is more significant than the earnings figures themselves. Prior to this, among the 34 Wall Street analysts covering Intel, 24 had issued a Hold rating, with an average target price of $55.33. At that time, Intel's stock price was already at $66—indicating that the majority of institutional assessments were not merely conservative but lagging behind reality. A quarterly report that significantly exceeded expectations did more than just boost the stock price; it exposed this gap. The 20% after-hours surge can be seen, to some extent, as the market executing a delayed pricing correction on behalf of these institutions.
**Exceeding Expectations Across the Board**
Revenue reached $13.6 billion, surpassing the analyst consensus of $12.4 billion by approximately 9.3%. Adjusted EPS was $0.29, compared to an expected $0.01. The midpoint of the Q2 revenue guidance is $14.3 billion, exceeding the consensus estimate of $13.1 billion. All three core metrics substantially outperformed expectations, and the degree of outperformance was systemic, not marginal—the implied volatility priced into the options market before the release was 9.3%, whereas the actual move was more than double that, suggesting even hedging institutions were caught off guard.
| Metric | Q1 2026 Actual | Analyst Expectation | Q1 2025 | | :--- | :--- | :--- | :--- | | Revenue | $13.6B | $12.4B | $12.7B | | Adjusted EPS | $0.29 | $0.01 | $0.18 | | Non-GAAP Gross Margin | 41.0% | ~39% | 39.2% | | Data Center & AI | $5.1B | — | $4.2B (+22%) | | Client Computing | $7.7B | — | $7.6B (+1%) | | Q2 Revenue Guide (Midpoint) | $14.3B | $13.1B | — |
The net loss under GAAP was $3.7 billion. The approximately $4 billion difference between non-GAAP and GAAP figures stems from stock-based compensation, depreciation, and restructuring charges. These represent the real costs associated with Intel's foundry transformation—they are not accounting noise, but neither are they new information for the market, as analysts had already factored them into their valuations.
**What Were the 24 Hold-Rated Institutions Waiting For?**
The consensus Hold rating was not primarily a valuation issue. Over the past two years, as Intel successively lost its process leadership and saw market share eroded by AMD, Wall Street's deepest skepticism centered on execution. The narrative presented sounded promising, but Intel had promised "this time will be different" too many times before. Thus, even with a clear demand logic from AI agents and a plausible story for CPU value reassessment in the inference layer, most institutions were waiting for concrete proof rather than betting on the narrative alone.
This earnings report provided precisely that proof—in the form of contracts. Intel announced it will manufacture chips for the Terafab factory, with customers including SpaceX, xAI, and Tesla. On the same day, a multi-year agreement was signed with Google, where Xeon CPUs will provide computing power for AI inference and other workloads on Google Cloud. The simultaneous announcement of these two contracts with the earnings release was not coincidental; it was management demonstrating in black and white that the discussed demand is not merely a trend observation but represents secured revenue. This is also the source of confidence behind the Q2 guidance—$13.8 to $14.8 billion, with a midpoint $1.2 billion above consensus—a commitment backed by existing orders.
**The Data Center Narrative Has Shifted**
Revenue for the Data Center and AI business reached $5.1 billion, a 22% year-over-year increase, with an operating margin of 31%. Prior to 2023, this segment was often viewed as a remnant after NVIDIA's dominance. However, the large-scale deployment of AI agent architectures has altered the demand structure. While GPUs are used for large model training, the demand for CPUs in inference—particularly for enterprise-focused, latency-sensitive, medium-scale inference workloads—was severely underestimated. As competition for GPU resources among cloud providers intensifies, they have begun procuring server CPUs in large quantities to handle these tasks, with Intel's Xeon being the most reliably supplied option for this scenario. Mentioned during the earnings call that multiple customers are "actively evaluating" the next-generation 18A process—the weight of this phrasing, not "interested" but "actively evaluating"—suggests the commercialization of the foundry business is progressing from the意向 stage to a more substantive phase.
In contrast, the Client Computing Group revenue was $7.7 billion, up just 1% year-over-year. The "AI PC" concept has been discussed for two years, and Intel has launched its Core Ultra series, but this upgrade cycle has yet to generate significant demand traction. This segment contributes over half of the company's revenue yet shows minimal growth—it is the least compelling part of this earnings report and represents a genuine risk. Should the AI demand cycle in the data center fluctuate, the CCG segment offers little buffer.
**Foundry: The Road Ahead**
Intel Foundry Services revenue was $5.4 billion, a significant year-over-year increase. The sequential decline was only $72 million. While this rate of decline might be described in analyst reports as "progress in line with expectations," a simple calculation shows that at this pace, reaching breakeven for IFS will be measured in years, not quarters. A core reason for a pre-earnings rating upgrade was more optimistic data assessment regarding the 14A node—if 18A mass production proceeds smoothly, the customer appeal of 14A should increase further. However, each milestone on this path remains a reason for Hold-rated institutions to maintain their观望 stance and a factor that genuine bulls must continually validate.
**Potential Chain Reaction**
The 24 Hold-rated institutions now face an awkward situation. With an average target price of $55.33 and the stock touching $80 after hours, this gap can no longer be justified by "maintaining caution." When a stock price exceeds the average target by over 40%, analysts must either revise their assessment or admit they have fallen behind. Historically, whenever such concentrated institutional lag in pricing occurs, subsequent batch upgrades to target prices themselves can become a catalyst for the stock: each new research report publication acts as a widely reported bullish signal. A pre-earnings Street-high target price of $95 was already issued; if more institutions follow with upward revisions, this figure could become a new anchor for the market.
Two developments warrant close monitoring. First, the progression of the 18A process from "active evaluation" to "contract signing" and finally to "volume production"—each concrete step on this timeline will directly impact the pace of IFS's loss reduction and Intel's attractiveness as a foundry for external customers. Second, the repositioning of those 24 Hold-rated institutions—not because their judgment dictates the stock's direction, but because their collective adjustment provides an observable signal. When the institutional consensus shifts from predominantly "Hold" to predominantly "Buy," it typically signifies a fundamental change in the market's qualitative assessment of a company.
This earnings report did not merely alter perceptions of Intel's quarterly performance; it changed the core question itself. For the past two years, the prevailing question was "Can Intel survive?" Following this report, that question has largely been answered. The new questions are: How well can it thrive, and how quickly?
