Research Finder
Find by Keyword
Intel Q1 2026: Is Terafab the Anchor Customer Foundry Has Been Waiting For?
Intel delivers 6th straight beat: Xeon demand surging, 18A yields outpacing targets, and Terafab inclusion reframing foundry trajectory towards 14A
04/24/2026
Key Highlights
- Intel reported Q1 2026 revenue of $13.6 billion, up 7% year-over-year, with non-GAAP EPS of $0.29 versus prior guidance of approximately breakeven.
- Data Center and AI (DCAI) revenue reached $5.1 billion, up 22% year-over-year, while Intel Foundry revenue grew to $5.4 billion, up 16% year-over-year.
- Intel Xeon 6 was selected as the host CPU for NVIDIA's DGX Rubin NVL8 systems, while Google committed to a multiyear Xeon deployment across its C4 and N4 instances.
- Management guided Q2 revenue of $13.8 billion to $14.8 billion, implying sequential momentum despite a modest gross margin step-down tied to Intel 18A ramp economics.
- The thesis we are testing is whether the AI era's CPU resurgence is structural or a temporary byproduct of supply-constrained accelerator economics and early generation agentic models.
The News
Intel reported first-quarter 2026 revenue of $13.6 billion, up 7% year-over-year, with non-GAAP earnings per share of $0.29 that exceeded prior guidance of approximately breakeven. Data Center and AI revenue grew 22% year-over-year to $5.1 billion, Intel Foundry revenue increased 16% to $5.4 billion, and Client Computing Group revenue rose 1% to $7.7 billion, with management citing AI-related businesses as roughly 60% of total revenue. CEO Lip-Bu Tan characterized the quarter as evidence that AI's shift toward inference and agentic workloads is driving demand for CPUs and advanced packaging, while CFO David Zinsner noted demand continued to outpace supply across all business units. For additional detail, see the Intel Q1 2026 earnings release.
Analyst Take
Our read on this quarter continues a theme we’ve been on for the last 4 quarters… Intel is no longer being evaluated on whether the turnaround is real. The market has (finally) moved past that question. The sixth consecutive quarter of beating expectations, reflected in a stock market surge on sentiment, suggests investors have already underwritten the thesis that Lip-Bu Tan's reset is working. One pure analyst point, can anybody imagine the Intel middle management layer circa 2024 being selected (even trumpeted) by Elon Musk and his team for Terafab?
What we are now watching is whether the composition of growth is durable, or whether Intel is surfing on a wave of temporary constraints in accelerator supply, and model variability driving variable workloads, pushing workloads back toward host CPUs. Here is the contrarian observation. The narrative coming out of this print, that CPUs are reasserting themselves in AI infrastructure, is possibly overstated. The ratio shift Tan described, with customers deploying server CPUs alongside accelerators in proportions moving back toward CPU, could equally reflect GPU allocation scarcity rather than architectural preference. It could also reflect the need for CPU flexibility as agentic workloads roll out in early generations prior to refinement. Intel's challenge is proving this is architecture, not arbitrage.
What Was Announced
Intel's first quarter financial results reflect revenue of $13.6 billion, non-GAAP gross margin of 41.0% (up 180 basis points year-over-year), and non-GAAP operating margin of 12.3% (up 690 basis points year-over-year). The segment composition is where the analytical texture lives. Client Computing Group delivered $7.7 billion with operating income of $2.5 billion, up 1% on revenue but with margin expansion tied to the uptake of Intel Core Ultra Series 3 on Intel 18A. Management presents the ramp as the fastest new product spool up in five years. Data Center and AI delivered $5.1 billion with $1.5 billion of operating income, roughly tripling the prior-year segment profit. Intel Foundry generated $5.4 billion in revenue but reported a $2.4 billion operating loss, reflecting the ongoing capital intensity of the 18A ramp. Cash from operations was $1.1 billion, offset by gross capital expenditures of approximately $5 billion, yielding adjusted free cash flow of negative $2 billion. GAAP results reflected a $(0.73) loss per share, driven by $4.1 billion of restructuring and other charges, which management attributed primarily to a goodwill impairment at the Mobileye reporting unit. The Q2 2026 outlook is architected around revenue of $13.8 to $14.8 billion, non-GAAP gross margin of approximately 39%, and non-GAAP EPS of $0.20, with the margin step-down designed to absorb a larger Intel 18A contribution early in its cost curve. Intel also repurchased the 49% minority interest in the Fab 34 Ireland joint investment entity, a move that aims to consolidate control over European capacity as advanced packaging backlog reaches what Zinsner characterized as billions of dollars per year. Also key in the period is the announcement that Intel joined the Terafab project as a strategic partner alongside SpaceX, xAI, and Tesla, a deal that positions Intel as a key technology and manufacturing partner for Elon Musk's one-terawatt annual compute ambition, with the consortium now constructing a dedicated facility in Austin.
Market Analysis
The competitive positioning here deserves careful reading. AMD has been steadily taking server CPU share since 2017 and passed Intel on server CPU revenue at the end of 2024, with EPYC Venice launching this year designed to extend that lead. Against that backdrop, Intel's 22% DCAI growth and the NVIDIA DGX Rubin NVL8 host CPU design win are genuinely meaningful. NVIDIA's decision to architect its flagship AI reference platform around Xeon 6, combined with NVIDIA's roughly 4% equity stake in Intel, signals a coordinated posture we have not previously observed between these two companies. The Google multiyear agreement for Xeon 6 in C4 and N4 instances, paired with co-development of custom ASIC infrastructure processing units, further validates Intel's positioning in the hyperscaler stack. There is a wide semiconductor industry analyst consensus that advanced packaging is emerging as a structurally constrained capability. Demand continues to outpace global capacity additions, which tends to support Zinsner's commentary that Intel's advanced packaging backlog reached billions of dollars per year in the quarter. The Terafab consortium structure, where Tesla and SpaceX appear to be financing an Intel-operated facility while branding it as a Musk enterprise venture, is an unusual financing architecture that reminds us more of anchor-tenant commercial real estate than traditional foundry supply agreements. Intel Foundry's external revenue of approximately $174 million remains modest relative to the $5.4 billion segment total, continuing to reflect heavy internal consumption as Intel makes its own products for sale. The SambaNova heterogeneous hardware blueprint, combining GPUs, SambaNova RDUs, and Xeon 6 as host and action CPUs, is the kind of partnership that aims to position Intel within alternative inference architectures beyond the NVIDIA-dominant training stack.
Looking Ahead
Based on what we are seeing now, the critical watch items for the next two quarters are the Intel 18A yield trajectory and the pace at which external foundry design commitments convert into revenue. Management indicated that 14A maturity and performance are outpacing 18A at a comparable point in the ramp. The companies evidence here is earlier design commitments expected to emerge in the second half of 2026 and expand into the first half of 2027. Operating expenses are likely to run above the prior $16 billion target due to inflationary pressures and targeted investments, and that is going to compress the operating leverage story somewhat. Our analysis suggests the Terafab anchor customer, combined with the NVIDIA and Google engagements, has fundamentally altered the foundry narrative from speculative to contingent on execution. The rising memory input costs Zinsner flagged for the second half deserve monitoring. If the CPU-alongside-accelerator ratio shift proves architectural rather than allocation-driven, Intel's thesis holds. If GPU supply loosens and the ratio swings back, the durability question reopens.
Stephen Sopko | Analyst-in-Residence – Semiconductors & Deep Tech
Stephen Sopko is an Analyst-in-Residence specializing in semiconductors and the deep technologies powering today’s innovation ecosystem. With decades of executive experience spanning Fortune 100, government, and startups, he provides actionable insights by connecting market trends and cutting-edge technologies to business outcomes.
Stephen’s expertise in analyzing the entire buyer’s journey, from technology acquisition to implementation, was refined during his tenure as co-founder and COO of Palisade Compliance, where he helped Fortune 500 clients optimize technology investments. His ability to identify opportunities at the intersection of semiconductors, emerging technologies, and enterprise needs makes him a sought-after advisor to stakeholders navigating complex decisions.