Intel's 2026 AI Infrastructure Growth Strategy and CapEx Plan

Intel’s AI spend and 2026 CapEx plan to support infrastructure growth

Based on Intel’s Q1 2026 earnings transcript, Intel’s approach to funding AI-driven infrastructure growth in 2026 centers on (1) sustaining CPU supply for AI workloads and (2) step-up investments in leading-edge manufacturing capacity and productivity—while keeping overall capital intensity roughly flat year-over-year.

---

1) What Intel is “spending for” in 2026: AI infrastructure demand anchored on CPUs

Intel frames the AI infrastructure opportunity as increasingly CPU-anchored—with CPUs acting as the orchestration/control plane for AI systems that still use accelerators for core compute. Management says the CPU is being re-asserted as a “foundation” of the AI stack and that customers are deploying server CPUs alongside accelerators with the CPU-to-accelerator ratio moving back toward CPU. ¹

Intel also links the AI infrastructure buildout to demand for leading Xeon server CPUs, including products based on Intel 3 (Xeon 6) and Intel 18A (Core Series 3), which are described as being in full volume ramp and “maximizing and optimizing” factory output to meet customer needs. ²³

Implication for AI spend: Intel’s AI-related investment priorities (including where it directs engineering and manufacturing capacity) are closely tied to scaling CPU supply and related platform capabilities (including advanced packaging and foundry offerings) rather than treating AI as purely an accelerator-only story. ¹²

---

2) 2026 CapEx: flat year-over-year, but with a shift toward capacity that grows wafer output

Intel explicitly guides that capital expenditures in 2026 are expected to be flat vs. last year (and higher than a prior “flat to down” expectation), to reflect increased capacity investment needs. ⁴

Intel further specifies how that CapEx is deployed:

• Expenditures are expected to be roughly equal across the year. ⁴
• Spending will be heavily weighted toward equipment that directly grows wafer outs to support growth “this year and next.” ⁴

So, while the headline CapEx level is flat, management describes the composition of CapEx as more growth-oriented—focused on production output and factory throughput rather than only general investment. ⁴

---

3) Why Intel expects higher OpEx while keeping CapEx flat: intentional step-up in leading-edge investments

In the quarter, Intel Foundry operating loss improved sequentially, but management also said results were “mostly offset” by increased operating expenses due to an intentional step-up in Intel 14A investments to support internal and external customer evaluations. ⁵

Looking ahead to 2026, Intel indicates:

• OpEx in 2026 was directionally targeted around $16 billion but is likely to be higher due to inflationary pressures, variable compensation, and targeted investments to capture opportunities. ⁴

Implication for AI infrastructure: Even if the capital spending remains flat, Intel expects to fund AI infrastructure growth with a combination of (a) production-capacity CapEx for wafer output and (b) higher operating investment—particularly around advanced-node readiness and customer evaluation support (e.g., Intel 14A). ⁵⁴

---

4) Execution priorities in 2026: fab productivity, yields, and managing 18A ramp headwinds

Intel’s plan for supporting infrastructure growth in 2026 includes operational execution goals that would allow it to meet AI-driven demand:

• Management says it will “push for gross margin expansion” as a top priority, supported by yield/throughput improvements across process nodes. ⁴
• It acknowledges Intel 18A is still early in its ramp, and rising input costs (especially memory) are headwinds in the second half of 2026. ⁴
• For 2026 factory supply, Intel expects the network to increase available supply in the third and fourth quarters at a more measured pace than previously anticipated, due to a base effect from stronger-than-expected first-half output. ⁴
• Management also ties 2026 execution to improving factory output to meet customer needs amid demand running ahead of supply (notably Xeon server CPUs). ²⁴

Implication for AI infrastructure: Intel’s ability to support AI infrastructure growth in 2026 depends on converting leading-edge ramp progress into volume with acceptable yields/throughput, while navigating cost inflation and early-ramp inefficiencies. ⁴²

---

5) Product ramp and customer demand signals underpin the infrastructure buildout

Intel reports demand strength for AI-driven businesses and explicitly connects it to CPU infrastructure growth:

• Management says demand continues to outpace supply for AI-driven businesses “especially for Xeon server CPUs.” ²
• In Q1, Intel reports AI-driven businesses are 60% of revenue and grew 40% year-over-year. ⁶
• It also notes sustained momentum in server CPU demand and “confidence in the sustained growth of CPUs, driven by the AI infrastructure buildout.” ⁷

For 2026 specifically, Intel also notes that earlier design commitments on Intel 14A are expected to emerge beginning in 2026 and expand into 2027. ⁸

How this affects AI spend/CapEx planning: Intel’s spending is aimed at ensuring the manufacturing and platform ecosystem can handle near-term AI demand (CPU-anchored systems and advanced packaging) while laying groundwork for next waves of designs entering 2026–2027. ²⁸

---

Bottom line

Intel’s planning for AI spend and 2026 CapEx to support infrastructure growth can be summarized as:

1. AI demand fulfillment is CPU-anchored: Intel expects the AI stack increasingly requires CPUs for orchestration/control plane, with improving CPU-to-accelerator ratios in customer deployments. ¹⁹
2. 2026 CapEx is flat year-over-year, but purposefully growth-focused: CapEx in 2026 is guided to be flat vs. last year and weighted toward wafer-out/throughput equipment to support committed demand. ⁴
3. Higher OpEx supports leading-edge readiness and adoption: Intel expects 2026 OpEx to be higher than a directional $16B due to inflation, compensation, and targeted investment (including Intel 14A evaluation support). ⁴⁵
4. Execution and ramp management are central: Intel emphasizes yield/throughput improvements while acknowledging 18A ramp is early and input cost pressures exist, shaping how quickly additional supply capacity can be made available. ⁴²

All of these elements together indicate Intel is funding AI infrastructure growth mainly through capacity and productivity investments at its fabs (plus supporting operating investment for advanced-node adoption), while keeping overall CapEx level flat. ⁴⁵