Intel Q2 Earnings Exceed Expectations, Market Reevaluates AI Infrastructure Position

• Intel's second-quarter revenue and non-GAAP profit exceeded expectations, and its third-quarter guidance also surpassed market forecasts.
• The focus of the market reassessment is whether data center growth and foundry improvements can offset higher capital expenditures.
• Underlying assets: INTC, AMD, TSM, TSLA, NVDA.

After Intel released its second-quarter 2026 earnings on July 23, its stock rose significantly in after-hours trading. The market reaction was triggered not by a single number, but by stronger-than-expected results across revenue, data center performance, foundry losses, and third-quarter guidance—all surpassing previously pessimistic outlooks.

The financial report showed that Intel's second-quarter revenue was $16.1 billion, a 25% year-over-year increase, with non-GAAP earnings per share of $0.42. The company expects third-quarter revenue to range from $15.8 billion to $16.8 billion and non-GAAP earnings per share of $0.38. Under GAAP, Intel reported a loss of $2.16 per share in the second quarter, with a net loss of approximately $11 billion, indicating that profitability has not yet fully turned positive.

For investors, Intel has been viewed over the past few years as a laggard in the AI era. The GPU narrative belongs to NVIDIA, advanced manufacturing to TSMC, and server CPUs continue to face pressure from AMD. This earnings report has prompted the market to reconsider: Can Intel still rejoin the AI infrastructure value chain?

This change lies in the fact that Intel is no longer relying solely on one business to drive its recovery. Growth in data center revenue, narrowing losses in manufacturing services, and improved operating cash flow are jointly supporting the post-market reassessment.

The first line is DCAI, or Data Center and AI business, which primarily includes products such as server CPUs. While GPUs handle the core acceleration for AI training and inference, data centers still require CPUs to manage general-purpose computing, scheduling, networking, and background tasks.

In the second quarter, Intel's DCAI revenue reached approximately $6.3 billion, a 59% year-over-year increase. While this figure alone does not prove that Intel has regained server market share from AMD, it clearly indicates that the expansion of AI data centers is driving renewed demand for server platforms.

The second line is Foundry, the chip manufacturing services business. Intel doesn't just make chips for itself—it also aims to manufacture chips for external customers. In the past, the market was most concerned that this business requires heavy investment, has slow customer validation, and unclear returns on advanced processes.

The financial report shows that Foundry's second-quarter revenue was approximately $5.8 billion, a 31% year-over-year increase, and its operating loss narrowed from $3.168 billion in the same period last year to $2.089 billion. The revenue for the Intel segment includes internal transactions; therefore, the $5.8 billion should not be directly interpreted as external foundry customer revenue, but rather as a signal of operational improvement.

Intel's position in AI cannot be judged solely by whether it has the strongest GPU. Large AI data centers purchase more than just accelerator cards—they also require CPUs, networking, memory, storage, and power systems. As cloud providers continue to expand their AI clusters, server platforms will enter new update cycles.

DCAI's 59% growth at least indicates that Intel has not been entirely marginalized in the era of AI infrastructure. It continues to generate incremental revenue from expanding server demand, prompting the market to reassess the company's position within the CPU cycle.

This layer of analysis must still retain its boundaries. DCAI’s strength may stem from expanded AI spending, but it could also include factors such as inventory restocking, product cycles, and low base effects. It demonstrates revenue growth, but alone, it cannot prove that the long-term competitive landscape has reversed.

DCAI is more like an entry ticket. Intel has returned to the AI data center table, but it will need to demonstrate over the next several quarters that this is not just a short-term rebound amplified by industry tailwinds.

The contract manufacturing business determines another layer of valuation. Whether Intel can transition from being a traditional IDM—designing and manufacturing its own chips—to becoming an external foundry option within the U.S. advanced manufacturing ecosystem is one of the key variables driving its market revaluation.

14A is the forward technological anchor here. It can be understood as Intel’s next-generation advanced manufacturing process, designed to meet the demand for higher-performance, lower-power chips. Management’s positive statements regarding advanced processes and customer collaboration have also led the market to associate it with the company’s foundry transformation.

However, 14A cannot be described as having undergone large-scale commercial validation. Intel’s second-quarter report explicitly highlights risk production for 18A-P and high-volume manufacturing for Panther Lake, while 14A remains primarily within the context of 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 large-scale external order.

Foundry's revenue growth and narrowed losses indicate improving operational metrics. Interest or collaboration leads from external customers suggest Intel has an opportunity to change the market perception that "no major clients are willing to bet on it." However, between these signals and scalable revenue, there remain challenges in yield, delivery, cost, and customer onboarding.

Higher capital expenditures are the most compelling aspect of this earnings report. According to Reuters, Intel has raised its 2026 capital expenditure forecast to over $20 billion and stated that 2027 capital expenditures will be significantly higher than in 2026. The more confident management becomes, the more investors need to do the math.

If demand for AI data centers persists, DCAI growth continues, and Foundry’s losses continue to narrow, higher capital expenditures will be interpreted by the market as securing capacity and access to advanced process nodes ahead of time. For a company aiming to regain leadership in manufacturing, without investment, it will be difficult to rebuild competitiveness.

However, if the growth is merely a temporary rebound, or if adoption by advanced process customers is slower than expected, capital expenditures will once again become a valuation pressure. Intel’s past most damaging factor to market confidence has been delays in advanced processes and insufficient returns on massive investments.

The cash flow metrics also need to be examined separately. Intel's operating cash flow for the second quarter was $7.006 billion, which remains positive after deducting gross capex; however, the company's adjusted free cash flow was -$8.419 billion, primarily due to items such as partner contributions. Simply stating that "free cash flow turned positive" obscures these methodological differences.

This earnings report is sufficient to overturn a previous assumption: Intel has not been completely left behind in the era of AI infrastructure. Its data center business has regained momentum, its foundry business has narrowed its losses, and its guidance for the third quarter provides the market with further reason to remain observant.

But the market must verify whether these improvements can endure through the capital expenditure cycle. Annual investments exceeding $20 billion indicate that Intel is betting not on minor tweaks, but on the simultaneous success of advanced manufacturing, AI servers, and external foundry services.

If future financial reports show that DCAI growth has not significantly slowed, Foundry losses continue to narrow, and advanced process customer leads convert into clearer volume production arrangements, Intel’s valuation anchor may shift from a “cyclical rebound stock” to an “AI infrastructure participant.”

Conversely, if cash continues to be consumed by capital expenditures and contract manufacturing clients remain stuck in the realm of narratives, this after-hours revaluation will be reclassified by the market as a earnings rebound amplified by the AI cycle. Whether Intel represents a turning point ultimately depends on whether growth can afford its manufacturing ambitions.