Earnings blowout sends stock surging after hours – has Intel finally turned the corner?

TL;DR

• Intel's second-quarter revenue and non-GAAP profit exceeded expectations, and its third-quarter guidance also surpassed market forecasts.
• The market's re-evaluation focuses on whether data center growth and foundry improvements can offset higher capital expenditures.
• Related tickers: INTC, AMD, TSM, TSLA, NVDA.

Following Intel's release of its Q2 2026 earnings report on July 23, the stock price rose notably in after-hours trading. The market's response was not triggered by a single metric, but by the fact that revenue, data center performance, foundry losses, and Q3 guidance all came in better than the pessimistic expectations that had been set.

According to the report, Intel's Q2 revenue was $16.1 billion, a year-over-year increase of 25%, with non-GAAP earnings per share of $0.42. The company forecasts Q3 revenue between $15.8 billion and $16.8 billion, and non-GAAP EPS of $0.38. On a GAAP basis, the company reported a loss per share of $2.16 for Q2, resulting in a net loss of approximately $11 billion. The improvement in profitability does not yet represent a complete turnaround.

For investors, Intel has been perceived as a laggard in the AI era over the past few years. The GPU narrative belongs to Nvidia, the advanced manufacturing narrative to TSMC, and its server CPUs face competitive pressure from AMD. This earnings report prompts the market to re-evaluate: can Intel still find a place within the AI infrastructure supply chain?

Three Lines of Business Drive Market Repricing

The change this time is that Intel is no longer relying on a single business segment to tell a recovery story. The after-hours revaluation was supported by three pillars simultaneously: growth in data center revenue, narrowing foundry losses, and improved operating cash flow.

The first line is DCAI, which stands for Data Center and AI business, primarily including server CPUs and other products. While the core acceleration for AI training and inference is handled by GPUs, data centers still require CPUs for general-purpose computing, scheduling, networking, and background tasks.

In Q2, Intel's DCAI revenue reached approximately $6.3 billion, up 59% year-over-year. This figure doesn't directly prove that Intel has regained server market share from AMD, but it strongly suggests that the expansion of AI data centers is also driving a refresh cycle for server platforms.

The second line is Foundry, the chip manufacturing business. Intel not only manufactures chips for itself but also aims to produce chips for external customers. The market's biggest concern has been the high investment required for this segment, slow customer qualification, and unclear returns on advanced process nodes.

The report shows that Foundry generated approximately $5.8 billion in Q2 revenue, up 31% year-over-year, with operating losses narrowing to $2.089 billion from $3.168 billion in the same period last year. Since Intel's segment revenue includes internal transactions, the $5.8 billion figure cannot be directly interpreted as revenue from external foundry customers; it is better viewed as a signal of improving operational efficiency.

DCAI Brings Intel Back into the AI Server Narrative

Intel's position in AI cannot be assessed solely based on whether it possesses the most powerful GPU. Large AI data centers purchase not only accelerator cards but also CPUs, networking equipment, memory, storage, and power systems. As long as cloud providers continue to build out AI clusters, server platforms will enter a new upgrade cycle.

The 59% growth in DCAI, at a minimum, indicates that Intel has not been completely marginalized by the AI infrastructure era. It can still generate incremental revenue from the expansion of server demand, leading the market to reassess the company's position within the CPU cycle.

However, this assessment must be tempered with caution. The strong performance of DCAI could stem from increased AI spending, but it might also include factors like inventory replenishment, product cycles, and a low base of comparison. It proves revenue is growing, but alone, it does not confirm a reversal in the long-term competitive landscape.

DCAI can be seen more as an entry ticket. Intel has secured a seat at the AI data center table, but it needs to demonstrate over subsequent quarters that this is not merely a short-term rebound amplified by favorable industry conditions.

Foundry Improvement Remains in a Verification Phase

The foundry business determines another layer of valuation. Whether Intel can transform from a traditional IDM (a company that designs and manufactures its own chips) into a viable external foundry option within the U.S. advanced manufacturing supply chain is a key variable for a potential market re-rating.

14A serves as the long-term technological anchor here. It can be understood as Intel's next-generation advanced manufacturing process, aimed at serving chip demands requiring higher performance and lower power consumption. Management's positive stance on advanced process nodes and customer partnerships has linked 14A with the foundry transformation narrative in the market's view.

However, 14A cannot yet be characterized as having undergone large-scale commercial validation. Intel's Q2 report more explicitly discusses risk production for 18A-P and high-volume manufacturing for Panther Lake. 14A still primarily appears in forward-looking statements and risk disclosures. At this stage, it is more appropriate to view it as a technological narrative anchor rather than a confirmed source of substantial external orders.

Foundry's revenue growth and narrowing losses indicate that operational metrics are improving. Expressions of interest or collaboration clues from external customers suggest that Intel has an opportunity to change the market's perception that it 'lacks major clients willing to commit.' However, the path from signals to scalable revenue generation still involves hurdles such as yield rates, delivery schedules, costs, and customer qualification.

Higher CapEx Increases Pressure on Cash Flow

Higher capital expenditure is the most tension-laden aspect of this earnings report. According to Reuters, Intel raised its CapEx guidance for 2026 to over $20 billion and stated that CapEx in 2027 will be significantly higher than in 2026. The more confident management appears, the more investors scrutinize the numbers.

If AI data center demand persists, DCAI growth continues, and foundry losses keep narrowing, higher CapEx could be interpreted by the market as a proactive move to secure production capacity and access to advanced process nodes. For a company aiming to regain manufacturing leadership, failing to invest would make it difficult to rebuild competitiveness.

However, if the growth proves to be a temporary rebound, or if customer adoption of advanced process nodes is slower than expected, the CapEx will once again become a drag on valuation. Historically, what has hurt investor confidence in Intel the most are delays in advanced process technology and insufficient returns on massive investments.

The cash flow metrics also require careful examination. Intel reported Q2 operating cash flow of $7.006 billion, which remained positive after deducting gross capex. However, the company's disclosed adjusted free cash flow was -$8.419 billion, primarily affected by items such as partner contributions. A simplistic statement like 'free cash flow has turned positive' would obscure these accounting differences.

Capital Returns Determine the Nature of the Turnaround

This earnings report is sufficient to change an old assessment: Intel has not been entirely left behind by the AI infrastructure era. Its data center business is re-accelerating, foundry losses are narrowing, and the Q3 guidance provides a reason for the market to continue monitoring the situation.

However, what the market needs to verify is whether these improvements can withstand the cycle of capital expenditure. An annual investment exceeding $20 billion indicates that Intel is not betting on minor adjustments, but rather that three lines – advanced manufacturing, AI servers, and external foundry – will all prove viable simultaneously.

If future earnings reports show that DCAI growth is not significantly decelerating, Foundry losses continue to narrow, and customer leads for advanced process nodes translate into clearer mass production plans, then Intel's valuation anchor might shift from a 'cyclical rebound stock' toward an 'AI infrastructure participant.'

Conversely, if cash continues to be consumed by CapEx while foundry customers remain a narrative, this after-hours revaluation could be reclassified by the market as a transient earnings bounce amplified by the AI cycle. Whether Intel represents a true turning point ultimately depends on whether its growth can sustain the cost of its manufacturing ambitions.