Applied Materials and Intel told investors on October 6, 2026 that they are deepening a partnership built around Applied’s EPIC Center in Silicon Valley, the chipmaking equipment giant’s newest research hub for next-generation transistors, interconnects, and advanced packaging. The announcement landed the same day Applied Materials posted $9.12 billion in fiscal third-quarter revenue, and it arrives five months after Applied named TSMC, Intel’s biggest foundry rival, as a founding partner of the same facility. That overlap is the real story: two competitors sharing R&D floor space at the company that sells both of them the tools to make chips.
For hardware watchers tracking the AI chip buildout, the deal is less about a single product and more about where the money and the research hours are going next. Intel Foundry posted a $2.1 billion operating loss last quarter. Intel’s own capital spending guidance for 2026 has climbed past $20 billion. And the company whose equipment every leading-edge fab depends on just said, in public, that Intel and TSMC will now work side by side at its newest lab.
What Applied Materials and Intel Actually Announced
The release, published via GlobeNewswire at 9:00 a.m. ET on October 6, 2026, describes a collaboration spanning next-generation transistors, interconnects, and advanced packaging technologies, built around two sites: Applied’s EPIC Center in Silicon Valley and Intel’s research campus in Hillsboro, Oregon. Applied Materials CEO Gary Dickerson framed it as an extension of a relationship that goes back decades, pointing to a long shared history of collaboration on materials engineering across the semiconductor roadmap. Intel CEO Lip-Bu Tan tied the expanded work directly to AI demand, citing the need for gains in performance, power efficiency, and packaging, and describing Applied as an important partner in getting new technology from the lab into high-volume manufacturing in the United States.
Notably absent from the release: a dollar figure. There is no disclosed investment amount, no percentage improvement target, and no mention of a specific Intel process node like 18A or the still-distant 14A. That omission matters for anyone trying to size this deal. It reads as a research and engineering commitment, not a funding round, and the companies worded it that way on purpose.
Inside the EPIC Center: From a $4 Billion Pledge to a $5 Billion Hub
Applied Materials first unveiled the EPIC Center, short for Equipment and Process Innovation and Commercialization, back in May 2023, pledging up to $4 billion in incremental capital investment over seven years. The original plan called for more than 180,000 square feet of cleanroom space, up to 2,000 new engineering jobs, and as many as 1,500 construction jobs during the build-out. By May 2026, Applied and TSMC were describing the completed project as a $5 billion facility, a sign that costs ran well ahead of the initial estimate, which happens often in cleanroom-heavy semiconductor construction.
The center is scheduled to be operationally ready sometime in 2026, according to Applied Materials, which positions Intel’s newly expanded role as one of the first major tenant partnerships to kick in right as the lights come on. For context on how capital-intensive this corner of the chip industry has become, see our coverage of TSMC’s own multibillion-dollar US fab expansion plans.
Why Intel Is Leaning Harder on Applied Materials Right Now
Timing explains a lot here. Intel Foundry, the division built to manufacture chips for outside customers as well as Intel’s own product lines, posted $5.8 billion in revenue and a $2.1 billion operating loss in its most recently reported quarter, an improvement of $348 million from the prior period but still a steep hole. Closing that gap requires Intel to win more external customers, and winning customers requires proving that its advanced nodes and packaging can compete with TSMC on schedule and yield, not just on paper.
Intel’s 18A node reached volume production in Arizona, with the company stating on its own Intel Foundry site that it has been running at volume there since December, a claim echoed in separate reporting. Getting from “node works” to “node wins outside customers” is where deeper materials and packaging collaboration pays off, which is likely why Naga Chandrasekaran, Intel Foundry’s chief technology and operations officer, said in the release that accelerating R&D to keep pace with AI infrastructure buildout is critical to delivering the manufacturing excellence Intel’s customers expect.
The Technology at Stake: Transistors, Interconnects, and Packaging
The release groups the work into three buckets: next-generation transistors, interconnect scaling, and advanced packaging. None of these are new problems in chipmaking, but all three have gotten harder and more expensive as AI chips pack more compute into tighter power and thermal envelopes.
Next-generation transistors and interconnect scaling
Applied and Intel’s teams are working on new materials, process technologies, and device architectures aimed at advanced logic nodes built for AI workloads. The release specifically calls out both front-end-of-line work, the transistor layer itself, and back-end-of-line work, the wiring and interconnects stacked above it, which together determine how fast and how efficiently a chip can move data internally. As nodes shrink, interconnect resistance becomes a bigger bottleneck than the transistors themselves, a problem the chip industry has been fighting since roughly the 10-nanometer era.
Advanced packaging and Foveros-based 3D stacking
The more interesting piece, arguably, is packaging. The release ties the collaboration to Foveros-based 3D stacking, Intel’s technology for layering chiplets vertically rather than spreading them across a flat substrate. Done well, 3D stacking raises interconnect density, improves power delivery, and helps manage heat, all of which matter for what the release calls high-performance compute platforms designed to accelerate emerging agentic AI workloads. That phrase is doing a lot of work. It signals that both companies see packaging, not just raw transistor density, as the next competitive battleground for AI silicon.
TSMC Is Already in the Room, and That Changes the Calculus
Here is the detail that makes this deal worth more than a routine partnership update: TSMC is already a founding partner of the EPIC Center, a status Applied Materials confirmed back in May 2026. That means Intel Foundry’s most direct competitor has had a seat at the same research table for months before Intel’s expanded role was announced. Applied Materials has effectively positioned itself as neutral ground, selling leading-edge process and packaging tools to whichever foundry is willing to co-develop them fastest.
For Intel, that’s an uncomfortable but pragmatic reality. Shared access to the same equipment ecosystem that serves TSMC is the price of staying on a competitive roadmap at all. For Applied Materials, it is simply good business: when the two largest logic foundries in the Western world are both racing to serve AI chip demand, selling tools and co-development time to both is how revenue grows without picking a side.
Intel Foundry’s Financial Reality
Numbers help explain why Intel needs this partnership to work. Intel Foundry’s $5.8 billion in quarterly revenue, detailed in earnings coverage from MarketBeat, is a fraction of what TSMC generates in a comparable period, and the $2.1 billion operating loss underscores how far the division remains from profitability even as it narrows the gap quarter over quarter. Intel has responded by raising its 2026 capital expenditure guidance to more than $20 billion company-wide, a figure driven, according to Intel’s own investor relations materials, by stronger customer-demand signals. That spending has to eventually translate into external foundry customers, not just internal Intel product volume, for the foundry business to turn a corner.
Our earlier reporting on AMD’s own warning that AI chip demand is outpacing supply into 2028 gives useful context here. Every chipmaker racing to meet AI demand is also racing to secure the equipment and packaging capacity to actually build the chips, and that competition for capacity is exactly what’s pulling Intel and Applied Materials closer together.
Applied Materials’ Earnings Backdrop
Applied Materials reported fiscal third-quarter 2026 total revenue of $9.12 billion, with its Semiconductor Systems segment, the unit that sells the equipment chipmakers need for leading-edge nodes, contributing $7.04 billion of that total. Those results landed on the same day as the Intel collaboration news, which is unlikely to be a coincidence. Equipment makers like Applied Materials benefit directly when foundries increase R&D spending and equipment orders, so pairing an earnings beat with a partnership announcement is a well-worn pattern in this industry.
For readers tracking the broader memory and equipment supply chain, our coverage of Micron’s record quarter and tightening memory supply and DRAM pricing now outpacing TSMC’s 2nm chip costs point to the same underlying pressure: AI demand is straining every layer of the chip supply chain at once, not just the foundries themselves.
Market Impact: What This Means for Chip Equipment Stocks
Applied Materials, Intel, and TSMC all trade as bellwethers for the broader semiconductor capital equipment cycle, and AMAT and INTC shares tend to move on announcements like this one even without a disclosed dollar figure attached. Investors read a deepened Intel-Applied partnership as a signal that Intel Foundry’s roadmap has enough institutional backing to be worth betting on, which matters for a division still bleeding money on an operating basis. For Applied Materials, a visible role serving both Intel and TSMC at the same facility reduces exposure to any single customer’s capex cycle, a diversification point analysts tend to reward.
It’s also worth watching how this plays against the broader chip equipment supply chain, including the silicon IP licensing deals covered in our report on Synopsys’ billion-dollar chip IP agreement with Amazon. The common thread across these deals is that AI compute demand has made the tooling and IP layers of chipmaking just as strategically important as the fabs themselves.
Competitive Comparison: Intel Foundry vs. TSMC vs. Samsung Foundry
The scale gap between Intel and TSMC remains the central fact of this entire story. TSMC raised its 2026 capital expenditure guidance to a range of $60 billion to $64 billion, up from an earlier estimate of roughly $56 billion, according to its Q2 2026 earnings call. Intel’s company-wide capex guidance, by contrast, sits at just above $20 billion. Samsung Foundry’s standalone financials are harder to pin down since Samsung Electronics reports foundry results folded into its broader semiconductor segment rather than as a standalone line item, so comparable capex or revenue figures for Samsung Foundry specifically aren’t independently verifiable from current public disclosures.
| Foundry | 2026 Capex Guidance | Leading-Edge Node Status | EPIC Center Role |
|---|---|---|---|
| Intel Foundry | >$20B (company-wide) | 18A in volume production (Arizona, since December) | Expanded partner, announced Oct. 6, 2026 |
| TSMC | $60B-$64B (raised from ~$56B) | Advanced nodes ramping, exact 2026 volume-production node not independently confirmed | Founding partner, confirmed May 11, 2026 |
| Samsung Foundry | Not separately disclosed from Samsung Electronics semiconductor segment | Not independently confirmed for 2026 | No confirmed role |
That capex gap, roughly three times larger at TSMC than at Intel on a company-wide basis, is the single clearest number explaining why Intel needs partnerships like this one to punch above its own spending weight. Applied Materials’ equipment and co-development time becomes a force multiplier when a company’s own R&D budget can’t match the leader’s.
Historical Context: A Decades-Long Partnership Reactivated
Applied Materials and Intel have worked together since well before the current AI boom, a point Dickerson emphasized directly in the announcement. What changed is the venue and the urgency. The EPIC Center itself traces back to a wave of US semiconductor investment announcements in 2023, a period the Semiconductor Industry Association has tracked as part of a broader reshoring push, when the original $4 billion, 2,000-job pledge was framed partly around bringing advanced manufacturing research back to Silicon Valley. Three years later, the same facility has grown into a $5 billion hub serving the two companies at the center of the AI chip race, Intel and TSMC, simultaneously. The throughline from 2023 to 2026 is a steady escalation: what began as a reshoring and jobs story has become an AI infrastructure story, with both the dollar figures and the stakes roughly doubling along the way. Readers following Intel’s parallel node roadmap can find more detail in our report on Intel’s Panther Lake laptop rollout, which runs on the same broader 18A process family this collaboration is meant to strengthen.
The Agentic AI Angle: Why Packaging Now Matters More Than Transistors
The release’s specific reference to agentic AI workloads is a tell. Agentic AI systems, the kind that chain together multiple model calls and tool invocations rather than answering a single prompt, place different demands on hardware than training runs do. They need sustained, high-throughput inference at scale, which pushes packaging, memory bandwidth, and power delivery to the front of the priority list ahead of raw transistor count. Foveros-based 3D stacking addresses that need directly by shortening the physical distance data has to travel between compute and memory tiles, cutting both latency and power draw.
That’s a meaningful shift from a few years ago, when leading-edge node announcements were almost entirely framed around transistor density and clock speed. Packaging has become the quieter but increasingly decisive battleground, and this partnership signals that both Applied Materials and Intel see it that way.
What Comes Next: Five Predictions
- Expect Intel to use early EPIC Center output as proof points in its pitch to external foundry customers, especially around packaging benchmarks tied to agentic AI inference rather than training performance.
- Watch for Applied Materials to announce additional founding or expanded partners at the EPIC Center over the next two to three quarters, given the facility is still ramping toward full operational readiness.
- Intel Foundry’s operating loss should keep narrowing gradually through 2027, but a return to profitability likely depends more on landing marquee external customers than on any single equipment partnership.
- TSMC’s capex lead over Intel, currently roughly three times larger company-wide, is unlikely to close meaningfully in the next two years, keeping Intel reliant on partnerships to offset its smaller R&D budget.
- Packaging technology, not raw node shrinkage, will likely dominate the next wave of foundry marketing from all three major players as agentic AI inference workloads become the primary sales pitch to enterprise customers.
Key Numbers Behind the Deal
| Metric | Figure | Source & Date |
|---|---|---|
| EPIC Center original pledge (2023) | Up to $4B over 7 years, 180,000+ sq ft, up to 2,000 jobs | Applied Materials release, May 22, 2023 |
| EPIC Center current valuation | $5 billion facility | Applied Materials & TSMC release, May 11, 2026 |
| Applied Materials Q3 FY2026 total revenue | $9.12 billion | Earnings coverage, Oct. 6, 2026 |
| Applied Materials Semiconductor Systems revenue | $7.04 billion | Earnings coverage, Oct. 6, 2026 |
| Intel Foundry quarterly revenue | $5.8 billion | Intel Q2 FY2026 earnings call, reported July 23, 2026 |
| Intel Foundry operating loss | $2.1 billion (improved $348M quarter over quarter) | Intel Q2 FY2026 earnings call, reported July 23, 2026 |
| Intel 2026 capex guidance | Raised to >$20 billion | Intel earnings call, reported Sept. 30, 2026 |
| TSMC 2026 capex guidance | Raised to $60B-$64B (from ~$56B) | TSMC Q2 2026 earnings call, reported Oct. 5, 2026 |
Why No Dollar Figure Was Disclosed for This Specific Deal
It’s worth sitting with the fact that the October 6 release contains zero financial figures tied to the newly expanded collaboration itself. That’s typical for research partnerships between an equipment vendor and its customer, as opposed to joint ventures, investment rounds, or supply agreements, which almost always carry a headline number. Readers should treat this as a scope-and-roadmap announcement rather than a spending commitment, and watch for follow-up disclosures, likely tied to Intel’s or Applied Materials’ future earnings calls, if either company decides to quantify the collaboration later.
Frequently Asked Questions
What is the Applied Materials EPIC Center?
The EPIC Center, short for Equipment and Process Innovation and Commercialization, is Applied Materials’ research and development hub in Silicon Valley. It was first announced in May 2023 with a pledge of up to $4 billion in investment over seven years and is designed to shorten the time it takes to move new chipmaking technology from early research into high-volume manufacturing.
Is this an investment or a joint venture between Intel and Applied Materials?
No. The October 6, 2026 announcement describes a research and engineering collaboration focused on transistors, interconnects, and packaging technology. No investment amount, equity stake, or joint venture structure was disclosed in the release.
Is TSMC also involved with the EPIC Center?
Yes. Applied Materials named TSMC a founding partner of the EPIC Center in a separate announcement on May 11, 2026, months before Intel’s expanded role was confirmed. Both foundries now work with Applied Materials at the same facility.
How much money is Intel Foundry losing right now?
Intel Foundry reported a $2.1 billion operating loss on $5.8 billion in revenue in its most recently reported quarter, an improvement of $348 million from the prior quarter, according to Intel’s Q2 fiscal 2026 earnings call.
What is Foveros-based 3D stacking?
Foveros is Intel’s packaging technology for stacking chiplets vertically rather than arranging them side by side on a flat substrate. The approach increases interconnect density, improves power delivery, and helps manage heat, which the companies say matters for AI compute platforms handling agentic AI workloads.
How does Intel’s chip spending compare to TSMC’s?
Intel’s 2026 capital expenditure guidance sits above $20 billion company-wide, while TSMC raised its 2026 guidance to a range of $60 billion to $64 billion. That means TSMC’s overall capex budget runs roughly three times larger than Intel’s.
When will the EPIC Center be fully operational?
Applied Materials has said the EPIC Center is scheduled to be operationally ready sometime in 2026, though an exact opening date has not been publicly disclosed.
Does this announcement mention Intel’s 18A or 14A process nodes specifically?
No. The release refers generally to advanced logic nodes and next-generation materials but does not name Intel 18A, 14A, or any other specific numbered process node.
Related
- TSMC Weighs Texas Fab Beyond $265B Arizona Bet
- Intel Panther Lake Laptops Hit OLED, CPUs Rise 10%
- AMD’s Su: AI Chip Demand Outpaces Supply Into 2028
- Micron Posts Record $54.2B, Tightens Memory Supply
- DRAM Now Costs 54% More Than TSMC 2nm Silicon
