```

After-hours on April 23rd, Intel surged 20%, with its stock price approaching $80, hitting a new high since the dot-com bubble.

But the increase itself is actually more thought-provoking than the financial report numbers. Before this, of the 34 Wall Street analysts following Intel, 24 rated it as Hold, with an average target price of 55.33. And at that time, Intel's stock price was already at 66—which means most institutions' judgments were not just conservative, but had already fallen behind reality.

A quarterly report far exceeding expectations didn't just push up the stock price; it also exposed this gap. This 20% post-market surge is, to some extent, the market completing a belated price correction for the institutions.

Comprehensively Exceeding Expectations

Revenue $13.6 billion, analyst expectation $12.4 billion, exceeding by about 9.3%. Adjusted EPS $0.29, expectation $0.01. Q2 revenue guidance midpoint $14.3 billion, expectation $13.1 billion.

All three core indicators significantly exceeded expectations—and not marginally, but systematically. The implied volatility for which options markets bet before the release was 9.3%, with the actual increase more than double—it means even hedging institutions were caught off guard.

Under GAAP, net loss was $3.7 billion. The roughly $4 billion difference between Non-GAAP and GAAP comes from stock-based compensation, depreciation, and restructuring costs. These are the real costs that Intel is bearing during its transition to foundry, not just accounting noise, but also not new information to the market—analysts have long priced this in.

What Are the 24 Hold Institutions Waiting For?

The Hold consensus is actually not mainly due to valuation.

In the past two years as Intel lost its process leadership and market share was eroded by AMD, Wall Street's biggest doubt about the company was execution: The story Lip-Bu Tan tells sounds good, but Intel has promised "this time will really work" too many times. So even if the demand logic for AI agents is clear, and the CPU's value in inference is being re-evaluated, most institutions are waiting for concrete proof, not just betting on a narrative.

This financial report provides exactly that proof—in the form of contracts. Intel announced it will manufacture chips for Musk's Terafab factory, with customers including SpaceX, xAI, and Tesla; on the same day, it signed a multi-year agreement with Google, where Xeon CPUs will power Google Cloud’s AI inference and other workloads.

The announcement of these two contracts on the same day as the quarterly report is no coincidence; it's management telling the market in black and white: The demand we speak of is not just a trend judgment, it's already secured revenue. That's also the confidence behind Q2 guidance—from $13.8B to $14.8B, the midpoint is $1.2B higher than consensus, a commitment backed by orders in hand.

The Logic of Data Centers Has Changed

Data Center and AI business revenue was $5.1 billion, a year-on-year increase of 22%, with an operating margin of 31%.

Before 2023, this division was seen as a post-NVIDIA-wreckage, but the large-scale deployment of AI agent architecture has changed the demand structure. Large-scale model training uses GPUs, but inference—especially for enterprise, latency-sensitive, medium-scale inference workloads—has been severely underestimated in terms of CPU demand.

At the same time as cloud vendors face white-hot competition for GPU resources, they have begun large-scale procurement of server CPUs for these tasks, and Intel's Xeon is currently the most reliable option for this scenario. Lip-Bu Tan mentioned in the conference call that multiple customers are "actively evaluating" the next-generation 18A process; the weight of this wording—not "interested," but "actively evaluating"—shows that the commercialization of foundry business is moving from intention to substantive stage.

In contrast, Client Computing business revenue was $7.7 billion, up 1% year over year. The "AI PC" concept on the PC side has been promoted for two years, and Intel has also launched its Core Ultra series, but this upgrade cycle has not yet brought genuine demand. This business contributes over half of Intel's revenues, but shows almost no growth—it's the least sexy part of this financial report, and also one of its greatest risks. If the data center AI demand cycle fluctuates, CCG has almost no buffer capacity.

Foundry: How Much Further Is There to Go?

Intel Foundry Services revenue $5.4B, up 16% year-on-year. $5.4B, up 16% $2.4B, the loss was narrowed from the previous quarter by just $72 million.

This pace of narrowing is usually described in investment bank analyst reports as "progress in line with expectations," but if you do the math—at this rate, it will take years rather than quarters for IFS to break even.

One of BNP Paribas' reasons for upgrading its rating before the report was more optimistic data evaluation for the 14A node (the successor to 18A)—if 18A is mass-produced smoothly, 14A’s customer appeal will rise further.

But every milestone on this road is a reason for Hold institutions to remain on the sidelines, as well as something true bulls must continue to verify.

Chain Reactions Ahead

The 24 Hold institutions are now in an awkward situation. The average target price is 55.33, after-hours price touched 55.33, after-hours price touched 80; this gap can no longer be justified by "staying cautious"—when the stock exceeds the target price by more than 40%, analysts must either revise their judgments or admit they're left behind.

Historically, whenever concentrated institutional lag in pricing occurs, the wave of subsequent target price upgrades also boosts momentum: every new report is a bullish signal reported by the media. HSBC had already given a street-high target of $95 before the earnings release; if more institutions follow with upgrades, that number could become the market’s new anchor.

Two developments are worth tracking. First is the timeline from "actively evaluating" 18A process to "signing contracts" and then to "mass production"—every concrete progress on this timeline will directly affect how fast IFS narrows its losses and impacts Intel’s foundry appeal for external customers.

Second is the repositioning of those 24 Hold institutions—not because their judgments decide the stock direction, but because their collective adjustment provides an observable signal—when consensus shifts from "mainly Hold" to "mainly Buy," it usually means the market’s fundamental view of a company has changed.

What this financial report changes is not Intel’s quarterly performance, but the nature of the question itself. For the past two years, the question has been "Can Intel survive?"; after this financial report, that's basically answered. The new question is—how well, and how soon.

Risk Warning and DisclaimerThe market carries risks and investment should be cautious. This article does not constitute personal investment advice, nor does it take into account the individual investment objectives, financial situation, or needs of any particular user. Users should consider whether any comments, opinions, or conclusions in this article are suitable for their particular circumstances. If you invest accordingly, you do so at your own risk. ```