The top 10 semiconductor foundries generated a record $53.49 billion in revenue in Q2 2026, up 11.5% quarter on quarter per TrendForce, as AI accelerators and HBM demand redraw the leading-edge capacity map. TSMC holds a record 72.5% of that revenue and is ramping 2nm steeply, with 2nm already 8% of its wafer revenue in Q3 2026. Samsung began trial 2nm production for Tesla at its Taylor, Texas fab in September 2026. Intel’s 18A reached high-volume production. SMIC is running at 93.7% utilization with China now 90% of its revenue.
Picking the wrong foundry partner in 2026 means missing your product window or watching your margins evaporate to AI accelerator allocations.
Here’s a breakdown of the 10 leading chip foundries, what they can fabricate today, who gets first call on capacity, and where the geopolitical pressure points sit. This page sits in our semiconductor manufacturing hub alongside guides to wafer inspection, reliability testing, and virtual metrology.
1. Taiwan Semiconductor Manufacturing Company (TSMC)
Best for AI, HPC, and flagship mobile SoCs at leading edge
TSMC isn’t just the largest among foundry semiconductor companies. As of Q2 2026 it holds a record 72.5% of top-10 foundry revenue (TrendForce) and roughly 90% of leading-edge production. If you’re designing an AI accelerator, a flagship phone SoC, or anything that needs N3/N2-class density, this is where the conversation starts and usually ends.
Process Node Leadership
TSMC moved N2 into volume production in late 2025 on gate-all-around nanosheet transistors, and the ramp is now steep: 2nm reached 8% of wafer revenue in Q3 2026, up from 3% in Q2, with 3nm at 33% and advanced nodes (7nm and below) at 79% of wafer revenue. The headline numbers versus N3E: 10 to 15% performance uplift or 25 to 30% power reduction, with more than 15% density gain. N2P with backside power delivery follows per TSMC’s 2026 roadmap, then A16 and the A13 variant aimed at AI and HPC workloads.
Yield data stays closely guarded, but the directional read is positive: initial N2 yields were reported in the 70 to 90% range at the start of volume production. The Q3 2026 financials back the ramp up: revenue of US$46.72 billion, up 41.2% year on year in US dollar terms, net income of NT$763.06 billion, up 68.7%, and gross margin of 66.4% even as the 2nm ramp dilutes margins. Q4 2026 revenue is guided to US$52.3 to 54.3 billion.
Capacity & Allocation Reality
Tier 1 anchors (Apple ~23–25% of revenue, Nvidia 6–10%, then Broadcom, Qualcomm, AMD) get earliest access and largest guaranteed allocations.
Tier 2 large fabless and IDMs (MediaTek, Marvell, Sony, Amazon) get substantial but negotiated allocations.
Tier 3 customers get pushed to second-wave variants (N3E/N3P instead of first-wave N3B, N5/N4 instead of N3).
As of mid-2026, N2 capacity was reported booked more than a year out, with customers reportedly locking capacity and prepaying months in advance; TSMC does not confirm allocation terms publicly. What it does confirm: a steep 2nm ramp into Q4 2026 on strong leading-edge demand.
Geopolitical Exposure
Taiwan produces more than 60% of global chips.
Taiwan’s government mandates that true cutting-edge R&D and manufacturing remain in Taiwan, so even as TSMC builds fabs in Arizona, Kumamoto, and Dresden, the tip of the spear stays on the island.
Per the fab plans released with its US expansion, Arizona Fab 2 brings 3nm production from 2028, and Fab 3 is slated for N2 and A16.
Anyone running multi-foundry strategies to hedge Taiwan risk is making a sensible call.
2. Samsung Electronics
Best for AI accelerators needing tight logic-memory integration
Samsung Foundry is the clear #2 in advanced nodes. It is technically aggressive on gate-all-around, structurally advantaged by sitting next to the world’s largest memory operation, and, as of Q2 2026, holds 5.9% of top-10 foundry revenue (TrendForce) with SMIC closing in behind it.
Process Node Leadership
Samsung beat TSMC to gate-all-around.
Its 3 nm GAA (MBCFET) entered production in 2022, claiming up to:
45–50% lower power
~23–30% higher performance
~16–35% area shrink versus 5 nm
Samsung’s 2nm GAA (SF2) is in mass production, and the 2026 proof point is concrete: Tesla’s AI5 chip taped out at Samsung in July 2026 on the SF2T variant, and trial 2nm AI5 wafer production started at the Taylor, Texas fab around September 15, 2026.
Yield Reality Is More Nuanced…
Samsung trailed TSMC on advanced-node yield for years, but industry reporting in 2026 puts its 2nm yield past the roughly 60% threshold that makes a high-volume customer viable. A customer like Tesla committing AI5 to SF2T is the strongest public evidence that the process is production-grade.
Capacity & Allocation Reality
Samsung’s “Shell-First” strategy pre-builds cleanroom shells so equipment can be installed quickly, and the company is investing more than $37 billion in central Texas across two advanced logic fabs and an R&D fab. The Taylor fab is attracting customers specifically looking beyond TSMC:
The Tesla anchor: a $16.5 billion deal signed in July 2025 puts AI5 (split with TSMC) and the follow-on AI6 on Samsung 2nm at Taylor, with mass-production verification targeted for end-2026 and volume supply ramping from 2027.
Tier 1 anchors (Qualcomm, Nvidia, Apple, Tesla, large AI chip partners) get earliest access to SF3/SF2 and preferential placement at Taylor on multi-year commitments.
Headroom: Samsung typically has more advanced-node headroom than TSMC because its share is smaller, which is exactly the second-source argument.
Geopolitical Exposure
Less concentrated than Taiwan-only TSMC, but still heavily exposed to Northeast Asian dynamics. Korea is deeply tied economically to China, security-tied to the US, and dependent on Japan for critical materials. The Taylor ramp is the hedge.
3. Intel Foundry
Best for sovereign US/EU manufacturing at advanced nodes
Intel’s foundry pivot is the most interesting development in the chip foundry companies landscape. Once synonymous with manufacturing leadership, then years of delays, Intel is now shipping on its most advanced node: 18A entered high-volume production at the start of 2026.
Process Node Leadership
Intel 18A combines GAA RibbonFET and PowerVia backside power delivery, and Intel got that combination to production before TSMC and Samsung. 18A entered high-volume production at the start of 2026:
Fab 52 in Arizona plus Oregon run 18A at a reported combined 30,000 wafers per month as of mid-2026.
Yields are reported stabilizing toward industry-standard levels expected in early 2027.
18A-P entered risk production in June 2026, offering a reported 9% performance gain or 18% power cut versus 18A.
The external-customer roadmap runs 18A-P, 18A-PT, then 14A.
14A is the stakes-raiser. Intel committed to completing 14A development in Q2 2026, with risk production reported for 2028 and high-volume manufacturing in 2029, and has said the node’s scale depends on landing significant external customers, with the commitment window running from H2 2026 into H1 2027.
Capacity & Allocation Reality
External foundry is still new for Intel, so allocation works differently than at TSMC or Samsung:
Anchor customers (US hyperscalers, top fabless, defense and government programs) get early access and genuine influence over the 18A-P and 14A ramps.
The US government holds a 9.9% equity stake in Intel, converted from CHIPS grants in August 2025, binding the foundry to US industrial policy.
External Customers Have Somewhat More Leverage Than At A Sold-Out Foundry…
Intel wants to win flagship AI and defense programs to validate its nodes, so it co-develops and allocates meaningful capacity against anchor wins.
Geopolitical Exposure
Intel’s structural advantage is geographic: fabs in Arizona, Oregon, New Mexico, Ireland, with Ohio under construction (now targeting around 2030 for its first New Albany fab), underwritten by the CHIPS Act and EU Chips Act. Less Taiwan and Korea dependence, more constrained on what can ship to China. For defense, aerospace, and national security buyers, this is a feature, not a bug.
4. GlobalFoundries
Best for mature-node specialty: RF, FD-SOI, automotive, Trusted Foundry
GlobalFoundries exited the 7nm race in 2018 to focus on “more-than-Moore” manufacturing: feature-rich nodes targeting applications where transistor density isn’t the constraint. That decision now looks prescient as automotive, RF, and power semiconductor demand grew faster than expected. In Q2 2026, GF posted $1.79 billion in revenue for a 3.2% share of the top-10 foundry market (TrendForce).
Process Node Leadership
GF’s portfolio runs from 180 nm down to 12 nm FinFET for select customers, with three platform pillars worth knowing:
FD-SOI (FDX family): 22FDX and 12FDX for low-power, RF, and mixed-signal designs in mobile, 5G, automotive, and IoT.
RF and analog: RF-SOI, SiGe, and RF CMOS strengthened by the IBM semiconductor acquisition – a leading position in RF silicon.
Embedded NVM and specialty: eMRAM on 22FDX, embedded flash, power management (BCD), plus emerging GaN and silicon photonics.
“Leading-edge” at GF means leadership in RF, FD-SOI, photonics, GaN, and heterogeneous integration on these nodes.
TSMC vs GlobalFoundries at mature nodes: TSMC offers mature nodes at scale but allocates them behind leading-edge priorities; GlobalFoundries builds its entire business on them. For RF-SOI, FD-SOI, and US Trusted Foundry requirements, GF is the specialist choice; for plain-vanilla 28nm logic at the lowest cost, TSMC and UMC usually win on price and capacity.
Capacity & Allocation Reality
GF runs a balanced multi-region network:
Malta (NY) and Essex Junction (VT) in the US
Dresden, Germany
Large Singapore campus
Announced investments include a €1.1 billion Dresden expansion through 2028 and roughly $16 billion in US expansion across facility modernization and AI-oriented silicon photonics R&D (figures as announced in 2025; re-check GF investor relations for updates). In Q2 2026, GF posted $1.79 billion in revenue for a 3.2% share of the top-10 foundry market (TrendForce).
Geopolitical Exposure
US-headquartered with fabs in the US, Germany, and Singapore, and no fabs in China or Taiwan. GF is the only commercial high-volume foundry with US “Trusted Foundry” designation, making it uniquely attractive for secure government and defense programs that need guaranteed US-based production.
5. United Microelectronics Corporation (UMC)
Best for predictable, mature-node automotive and industrial supply
UMC is the stable middle of the foundry semiconductor market: Taiwan-based, focused on mature and mid-range nodes with strong automotive and industrial positioning, and running hot. In Q3 2026 its capacity utilization hit 92% and operating income rose 24.7% quarter on quarter, on revenue of NT$74.24 billion (US$2.32 billion).
Process Node Leadership
UMC’s specialty leadership centers on 22/28nm, used widely in automotive, industrial, networking, display, and IoT, and the 22nm business keeps setting records: 18.3% of Q3 2026 wafer sales, the highest share yet. The 14nm FinFET node delivers about 50 to 60% power efficiency improvement versus 28nm.
The Big Strategic Move Is The Intel Collaboration:
A jointly developed 12nm FinFET process running in Intel’s Ocotillo fab in Arizona: process design kits land in late 2026, customer tape-outs and pilot production are expected in 2027, and meaningful volume follows in 2028, per UMC’s 2026 updates.
This gives UMC customers access to a Western FinFET-class platform for mobile, communications infrastructure, and networking, a real differentiator for buyers wanting to hedge Taiwan exposure without going leading-edge.
Capacity & Allocation Reality
UMC raised its 2026 cash capital spending to $2 billion and in October 2026 approved an additional expansion budget of roughly $3.2 billion to pull its buildout forward against customer product plans.
Q3 2026 utilization hit 92%, with operating income up 24.7% quarter on quarter.
Key customers include Texas Instruments, MediaTek, and Realtek, plus a long tail of automotive, industrial, and networking clients.
The Customer Dynamics Differ From TSMC:
Capacity is generally available but contract-driven, with high priority for long-lifecycle automotive, industrial, and networking programs rather than speculative one-off tape-outs. Once a design qualifies on a UMC process, customers tend to stay for many years.
Geopolitical Exposure
Taiwan-based with the associated risk, but UMC runs an unusually structured supply-chain risk monitoring platform tracking critical materials, sourcing regions, and geopolitical events. The Intel 12nm partnership adds Western capacity for select products. A magnitude 7.4 quake in eastern Taiwan in 2024 temporarily halted production at both TSMC and UMC fabs, a reminder of the structural exposure.
6. Semiconductor Manufacturing International Corporation (SMIC)
Best for China-centric designs requiring domestic fabrication
SMIC is China’s flagship foundry and the center of Beijing’s semiconductor self-sufficiency push. It is strong and expanding at mature nodes, pushing constrained 7nm-class capacity without EUV, and increasingly a domestic story: China accounted for 90.2% of SMIC’s revenue in Q2 2026, up from 84.1% a year earlier.
Process Node Leadership
SMIC offers foundry services from 350nm down to 7nm-class technologies, with most commercial volume at 28nm and above.
The N+1 process:
improves performance by about 20%
cuts power by 57%
reduces logic area by 63% versus 14nm
Advanced-node capacity is tiny relative to demand and heavily directed:
SMIC became a key foundry for Huawei and other sanctioned Chinese firms after TSMC stopped taking their orders.
Advanced output is directed toward domestic priorities, with Huawei the most prominent beneficiary.
Capacity & Allocation Reality
Q2 2026 was SMIC’s first quarter above $3 billion:
Revenue of $3.006 billion, up 20% quarter on quarter and 36.1% year on year.
Utilization at 93.7%, up from 93.1% in Q1 2026.
About 8,000 wafers per month of 12-inch capacity added in the quarter, bringing monthly capacity to about 1.1 million 8-inch-equivalent wafers.
China now 90.2% of revenue, up from 84.1% a year earlier.
Q3 2026 guided up another 2 to 4% at a 26 to 28% gross margin, with prices raised on the most in-demand capacity.
AI-adjacent demand (power management, interface, and supporting chips around AI systems) is what’s filling the fabs.
Geopolitical Exposure
SMIC was added to the US Entity List in 2020 with expanded controls since. Export restrictions limit access to advanced equipment (notably EUV), complicate servicing for existing tools, and stretch node migration timelines. For US, EU, or Japan-based fabless companies, SMIC carries elevated regulatory and reputational risk. For China-centric buyers, it is increasingly the only realistic local option.
7. Texas Instruments
Best for long-lifecycle analog and embedded for industrial and automotive
TI isn’t a foundry in the TSMC sense. It is a large analog and embedded IDM running 45 to 130nm-class fabs at scale, primarily for its own products. Worth understanding the distinction before evaluating.
Process Node Leadership
TI exited digital logic process development at 45nm, choosing to stop internal node scaling and rely on external pure-plays for 32nm and below where needed. Today TI focuses its internal manufacturing on analog and embedded products using 45 to 130nm processes on 300mm wafers, explicitly optimized for analog and embedded rather than chasing 7nm or 5nm.
Capacity & Allocation Reality
TI is midway through roughly $60 billion in US-centric expansion across seven fabs at three sites: Sherman and Richardson in Texas plus Lehi in Utah, all focused on analog and embedded 300mm wafers.
Geopolitical Exposure
Heavily US-centric expansion, with strong reshoring momentum.
Still exposed to US-China trade tensions, export controls, and global materials supply.
Compliance with EAR and ITAR limits what TI can ship to certain foreign customers or for certain applications.
8. Infineon Technologies
Best for automotive power, SiC, GaN, and security MCUs
Infineon is a vertically integrated power, automotive, and security IDM with its own fabs and specialty processes, so it is not a general-purpose logic foundry. Worth knowing what you’re buying: product supplier with selective manufacturing partnerships, not open foundry access.
Process Node Leadership
Core strengths sit in:
power semiconductors (MOSFETs, IGBTs, SiC, GaN)
automotive MCUs
security controllers
IoT/industrial MCUs
On CMOS logic, Infineon uses 40nm and 28nm nodes for security controllers and automotive/IoT MCUs.
The Wide-Bandgap Leadership Is The Real Differentiator:
Infineon was the first semiconductor manufacturer to develop 300mm GaN power wafer technology on existing high-volume infrastructure, significantly increasing potential output per wafer. SiC capacity gets first call on capital allocation alongside strategic fabs and R&D.
Capacity & Allocation Reality
Manufacturing is centered in Europe (Germany and Austria), with strategic 300mm power and wide-bandgap investments there. Capital allocation prioritizes strategic fabs, SiC capacity, and R&D for automotive, industrial power, and renewables.
Geopolitical Exposure
EU-anchored with global exposure.
Benefits from the European Chips Act.
Exposed to US-China and EU-China export controls.
9. NXP Semiconductors
Best for automotive platforms spanning mature to advanced nodes
NXP is a fab-light hybrid IDM: it designs mostly automotive and embedded chips and uses a mix of in-house mature fabs and outsourced advanced nodes (TSMC, GlobalFoundries, joint ventures). Not an open foundry, but a platform supplier with smart manufacturing partnerships.
Process Node Leadership
NXP uses in-house fabs for analog and mixed-signal and outsources advanced digital logic to foundry partners. The roadmap talk became product in 2026: at CES 2026, NXP unveiled the S32N7 super-integration processor series on a 5nm foundation, following the 5nm S32N55, putting 5nm automotive silicon in the market, with 3nm automotive parts on the roadmap through foundry partners.
Capacity & Allocation Reality
NXP holds a top-three global position in automotive semiconductors.
Capacity flows to its own portfolio first: automotive (MCUs, radar, domain controllers), industrial/IoT, mobile, and secure ID/payment products.
The Dresden JV pre-allocates 16/12nm capacity for NXP and partner automotive designs.
You can’t book NXP fab slots as a fabless customer. You become an NXP chip customer or, rarely, a JV or technology partner.
Geopolitical Exposure
Netherlands-headquartered, operating globally with manufacturing and suppliers primarily in Europe, Japan, and Taiwan.
Directly exposed to US-China and EU-China trade tensions.
The Dresden JV and Singapore VSMC fab are explicit attempts to secure localized, geopolitically diverse capacity reducing single-fab dependency.
10. STMicroelectronics
Best for European automotive, industrial, and FD-SOI MCUs
ST is an IDM focused on automotive, industrial, power, and MCUs, with European fabs. Like NXP and Infineon, it is a product-centric manufacturer, not an open leading-edge logic foundry.
Process Node Leadership
ST runs a wide range of mature and mid-range CMOS nodes (130/90/65/55/40/28nm) with digital, mixed-signal, and RF options for analog, power, and microcontroller products.
The Headline Development:
The STM32V8, launched in November 2025, is the industry’s first 18nm microcontroller: an Arm Cortex-M85 at 800MHz with 4MB of embedded phase-change memory (ePCM), built on ST’s 18nm FD-SOI process at its 300mm Crolles fab in France, with broader availability from Q1 2026. SpaceX selected it for the Starlink constellation.
The 18nm FD-SOI + ePCM process delivers around 50% better performance-per-power than ST’s 40nm embedded NVM node and 2.5x higher NVM density. ST states it is the only sub-20nm process supporting 3V operation, enabling integrated analog (power management, reset, clock, ADC) for robust industrial and automotive environments.
Capacity & Allocation Reality
ST is aggressively expanding 300 mm capacity in Europe at Crolles (France) and Agrate (Italy) to strengthen scale.
As an IDM, capacity goes to ST’s own products serving automotive, industrial, personal electronics, and communications equipment.
Historically during capacity crunches, ST has prioritized strategic partners and top customers.
Geopolitical Exposure
EU-headquartered (Geneva, with major operations in France and Italy) with a diversified supply chain across Europe, Taiwan, Japan, and the US.
Comparison: Top 10 Semiconductor Foundries 2026
Foundry
Leading-Edge Logic (≤3 nm)
Open Foundry Access
Multi-Region Production
Trusted/Sovereign Manufacturing
Specialty Nodes
TSMC
✔️
✔️
⚠️
❌
⚠️
Samsung
✔️
✔️
⚠️
❌
⚠️
Intel Foundry
✔️
⚠️
✔️
✔️
⚠️
GlobalFoundries
❌
✔️
✔️
✔️
✔️
UMC
❌
✔️
⚠️
⚠️
⚠️
SMIC
⚠️
✔️
❌
❌
⚠️
Texas Instruments
❌
❌
⚠️
✔️
✔️
Infineon
❌
❌
⚠️
⚠️
✔️
NXP
❌
❌
⚠️
⚠️
⚠️
STMicroelectronics
❌
❌
⚠️
⚠️
✔️
Legend:
✔️ Full ⚠️ Partial / limited ❌ Not supported
Semiconductor Foundry Market Share, Q2 2026
The top 10 semiconductor foundries earned a record $53.49 billion in Q2 2026, up 11.5% quarter on quarter, per TrendForce’s September 2026 ranking. The five largest by revenue share:
Rank
Foundry
Q2 2026 Revenue
Market Share
1
TSMC
~$40.2B
72.5% (record)
2
Samsung Foundry
$3.26B
5.9%
3
SMIC
>$3.0B
5.4%
4
UMC
~$2.1B
3.9%
5
GlobalFoundries
$1.79B
3.2%
Source: TrendForce, September 9, 2026. The notable movement: SMIC grew 20% in the quarter and is closing on Samsung for second place, while TSMC’s share set another record.
Watchlist for 2027
Two names sit outside this top 10 but belong on a 2027 shortlist.
Rapidus is Japan’s state-backed 2nm bet. Its IIM-1 pilot line in Chitose, Hokkaido has run since April 2025, a 2nm gate-all-around prototype built on IBM’s nanosheet architecture reached its expected electrical characteristics in July 2025, and the process design kit reached early customers in Q1 2026. Mass production is targeted for 2027, the company is reportedly in talks with more than 60 potential customers with IBM, Tenstorrent, and Broadcom named as leading prospects, and 1.4nm development starts in 2026. If the 2027 ramp lands, it becomes the first new leading-edge entrant in decades.
Vanguard International Semiconductor (VIS), a Taiwan-based mature-node specialist, re-entered eighth place in TrendForce’s Q2 2026 ranking on power management IC demand, a reminder that the mature-node tier below this list keeps consolidating around a few reliable names.
Which Foundries Have the Best Yield Performance in 2026?
No leading foundry publishes official yield data, so the honest comparison is reported maturity:
TSMC: 2nm (N2) entered volume production in late 2025 with initial yields reported in the 70 to 90% range, the strongest start of the three leading-edge players.
Samsung: 2nm (SF2) yield is reported in 2026 to have crossed the roughly 60% threshold where high-volume customers become viable, and Tesla committing its AI5 chip to SF2T is the clearest public evidence.
Intel: 18A is reported stabilizing toward industry-standard yields expected in early 2027, with defect density approaching the mature range after its first year of volume production.
The practical read for buyers: TSMC for lowest yield risk today, Samsung and Intel for leverage and headroom, with the gap narrowing through 2027.
Leading European Foundries for Mature-Node Mixed-Signal
European mature-node capacity centers on four names:
GlobalFoundries Dresden is the largest open-access option, with 22FDX FD-SOI and RF-SOI aimed exactly at mixed-signal, RF, and automotive work.
STMicroelectronics runs 300mm fabs at Crolles (France) and Agrate (Italy) across 130 to 28nm mixed-signal nodes, now extended by its 18nm FD-SOI process, though as an IDM its capacity serves its own products first.
Infineon manufactures power and mixed-signal in Dresden and Villach as an IDM.
X-FAB, headquartered in Germany, is the European pure-play specialty foundry for analog and mixed-signal CMOS, the default answer when a design needs open foundry access to European mature nodes below GF’s scale.
For hyperscaler-scale AI accelerators, the 2026 allocation math is simple: TSMC N2/N3 capacity is committed to anchor customers first, so new large programs either pay for priority, split volume across TSMC and Samsung (the Tesla AI5 model), or anchor Intel’s 18A-P/14A ramp in exchange for influence over it. Dual-foundry designs cost more up front and remove the single point of failure that a sold-out node represents.
How Do Semiconductor Foundries Compare on Sustainability?
Foundries do not publish comparable third-party sustainability ratings, so the practical comparison is public commitments and annual ESG reports. UMC, for example, states a net-zero emissions target by 2050 and runs supplier decarbonization programs with partners including Infineon, per its 2026 Q1 reporting. Buyers with Scope 3 reporting obligations should request each foundry’s latest ESG report during qualification rather than relying on rating aggregators.
Where AI Visual Inspection Fits In Semiconductor Manufacturing
Whichever foundry partner you choose, defect detection at submicron scale determines yield economics, and yield is what separates the foundries above. Averroes ai plugs into existing inspection equipment, feeding AI defect classification and segmentation directly into your process control loop, no new hardware required. Teams tracking fab output alongside partner selection can pair this page with our guide to manufacturing yield optimization.
99%+ classification accuracy with near-zero false positives
40 to 60% more submicron defects caught versus rule-based inspection
20 to 40 images per defect class to train and deploy in hours
WatchDog anomaly detection flags novel defects rule-based tools miss
On-premise or air-gapped deployment for fab-grade security and IP protection
The production evidence: in photomask manufacturing, Averroes sustained 99.993% classification accuracy across 1.4 million production images over 13 months at Tekscend Photomask, with defect review running 8x faster than the manual baseline.
Ready To Stop Losing Wafers To Escapes?
99%+ defect detection accuracy with near-zero false positives.
Semiconductor Foundries FAQs
What is the difference between a foundry and an IDM?
The difference between a foundry and an IDM comes down to who designs the chips. A pure-play foundry like TSMC fabricates chips designed by other companies (fabless customers). An IDM like Intel, TI, or Infineon designs and manufactures its own chips. Intel and Samsung run hybrid models where they do both.
Which semiconductor foundries can manufacture 2nm chips?
The semiconductor foundries in volume production at 2nm in 2026 are TSMC (N2, 8% of its wafer revenue by Q3 2026) and Samsung (SF2, with trial 2nm production also running at its Taylor, Texas fab). Intel’s 18A is a 2nm-class node with GAA and backside power delivery in high-volume production, offered externally through its 18A-P variant. Japan’s Rapidus targets 2nm mass production in 2027.
Why is TSMC so dominant in the semiconductor foundry market?
TSMC’s dominance in the semiconductor foundry market comes from three compounding advantages: process leadership at leading-edge nodes, a deep ecosystem of EDA tools and IP partners, and trust built over decades with fabless customers like Apple, Nvidia, and AMD. As of Q2 2026 it holds a record 72.5% of top-10 foundry revenue and roughly 90% of leading-edge production.
Who are TSMC’s main competitors?
TSMC’s main competitors are Samsung Foundry (5.9% share in Q2 2026, first to 2nm GAA mass production), Intel Foundry (18A in high-volume production, the sovereign US option), and SMIC (5.4% share, China’s domestic champion). At mature nodes, UMC and GlobalFoundries compete on specialty processes. Japan’s Rapidus becomes a leading-edge challenger if its 2027 2nm ramp lands.
Which foundry has the best yield in 2026?
No foundry publishes official yield data, but reported figures put TSMC ahead: initial 2nm yields of 70 to 90% at volume start in late 2025, versus Samsung’s 2nm reported crossing the roughly 60% viability threshold in 2026 and Intel’s 18A reported reaching industry-standard yields by early 2027. Mature-node yields at UMC and GlobalFoundries are stable and rarely the deciding factor.
How long does it take to qualify a new semiconductor foundry partner?
Qualifying a new semiconductor foundry partner typically takes 12 to 18 months for a new node tape-out and significantly longer for automotive-grade qualification. Advanced-node capacity also has to be reserved well ahead: as of 2026, leading-edge allocations are reportedly booked a year or more out, with prepayment common for strategic programs.
Conclusion
Ten semiconductor foundries, ten different bets on what matters most in 2026. TSMC owns leading-edge AI and HPC but charges scarcity premiums and concentrates Taiwan risk. Samsung delivers the credible #2 option with tight memory integration and US 2nm capacity now running trial wafers in Texas. Intel Foundry is the sovereign manufacturing play that finally has a high-volume node to back it. GlobalFoundries and UMC anchor the mature-node and specialty market for buyers who don’t need 3nm to win. SMIC serves China-centric designs at high regulatory cost. TI, Infineon, NXP, and STMicroelectronics are IDMs to buy from, not tape out at.
Whichever foundry partner you pick, yield economics decide whether your design ships on margin or burns through reserves. Averroes catches 40 to 60% more submicron defects on your existing inspection equipment. Book a free demo to see how AI visual inspection changes your foundry yield math.
![Top 10 Semiconductor Foundries [2026]](/_next/image?url=https%3A%2F%2F303922.eu15.myftpupload.com%2Fwp-content%2Fuploads%2F2024%2F11%2FTop-10-Semiconductor-Foundries.jpg&w=3840&q=75)
The top 10 semiconductor foundries generated a record $53.49 billion in revenue in Q2 2026, up 11.5% quarter on quarter per TrendForce, as AI accelerators and HBM demand redraw the leading-edge capacity map. TSMC holds a record 72.5% of that revenue and is ramping 2nm steeply, with 2nm already 8% of its wafer revenue in Q3 2026. Samsung began trial 2nm production for Tesla at its Taylor, Texas fab in September 2026. Intel’s 18A reached high-volume production. SMIC is running at 93.7% utilization with China now 90% of its revenue.
Picking the wrong foundry partner in 2026 means missing your product window or watching your margins evaporate to AI accelerator allocations.
Here’s a breakdown of the 10 leading chip foundries, what they can fabricate today, who gets first call on capacity, and where the geopolitical pressure points sit. This page sits in our semiconductor manufacturing hub alongside guides to wafer inspection, reliability testing, and virtual metrology.
Our Top 3 Picks
Best for AI and HPC at Leading Edge
TSMC
VIEW NOWBest for Western-Aligned Sovereign Manufacturing
Intel Foundry
VIEW NOWBest for Mature-Node Specialty (RF, FD-SOI, Trusted Foundry)
GlobalFoundries
VIEW NOW1. Taiwan Semiconductor Manufacturing Company (TSMC)
Best for AI, HPC, and flagship mobile SoCs at leading edge
TSMC isn’t just the largest among foundry semiconductor companies. As of Q2 2026 it holds a record 72.5% of top-10 foundry revenue (TrendForce) and roughly 90% of leading-edge production. If you’re designing an AI accelerator, a flagship phone SoC, or anything that needs N3/N2-class density, this is where the conversation starts and usually ends.
Process Node Leadership
TSMC moved N2 into volume production in late 2025 on gate-all-around nanosheet transistors, and the ramp is now steep: 2nm reached 8% of wafer revenue in Q3 2026, up from 3% in Q2, with 3nm at 33% and advanced nodes (7nm and below) at 79% of wafer revenue. The headline numbers versus N3E: 10 to 15% performance uplift or 25 to 30% power reduction, with more than 15% density gain. N2P with backside power delivery follows per TSMC’s 2026 roadmap, then A16 and the A13 variant aimed at AI and HPC workloads.
Yield data stays closely guarded, but the directional read is positive: initial N2 yields were reported in the 70 to 90% range at the start of volume production. The Q3 2026 financials back the ramp up: revenue of US$46.72 billion, up 41.2% year on year in US dollar terms, net income of NT$763.06 billion, up 68.7%, and gross margin of 66.4% even as the 2nm ramp dilutes margins. Q4 2026 revenue is guided to US$52.3 to 54.3 billion.
Capacity & Allocation Reality
As of mid-2026, N2 capacity was reported booked more than a year out, with customers reportedly locking capacity and prepaying months in advance; TSMC does not confirm allocation terms publicly. What it does confirm: a steep 2nm ramp into Q4 2026 on strong leading-edge demand.
Geopolitical Exposure
Taiwan produces more than 60% of global chips.
Taiwan’s government mandates that true cutting-edge R&D and manufacturing remain in Taiwan, so even as TSMC builds fabs in Arizona, Kumamoto, and Dresden, the tip of the spear stays on the island.
Per the fab plans released with its US expansion, Arizona Fab 2 brings 3nm production from 2028, and Fab 3 is slated for N2 and A16.
Anyone running multi-foundry strategies to hedge Taiwan risk is making a sensible call.
2. Samsung Electronics
Best for AI accelerators needing tight logic-memory integration
Samsung Foundry is the clear #2 in advanced nodes. It is technically aggressive on gate-all-around, structurally advantaged by sitting next to the world’s largest memory operation, and, as of Q2 2026, holds 5.9% of top-10 foundry revenue (TrendForce) with SMIC closing in behind it.
Process Node Leadership
Samsung beat TSMC to gate-all-around.
Its 3 nm GAA (MBCFET) entered production in 2022, claiming up to:
Samsung’s 2nm GAA (SF2) is in mass production, and the 2026 proof point is concrete: Tesla’s AI5 chip taped out at Samsung in July 2026 on the SF2T variant, and trial 2nm AI5 wafer production started at the Taylor, Texas fab around September 15, 2026.
Yield Reality Is More Nuanced…
Samsung trailed TSMC on advanced-node yield for years, but industry reporting in 2026 puts its 2nm yield past the roughly 60% threshold that makes a high-volume customer viable. A customer like Tesla committing AI5 to SF2T is the strongest public evidence that the process is production-grade.
Capacity & Allocation Reality
Samsung’s “Shell-First” strategy pre-builds cleanroom shells so equipment can be installed quickly, and the company is investing more than $37 billion in central Texas across two advanced logic fabs and an R&D fab. The Taylor fab is attracting customers specifically looking beyond TSMC:
Geopolitical Exposure
Less concentrated than Taiwan-only TSMC, but still heavily exposed to Northeast Asian dynamics. Korea is deeply tied economically to China, security-tied to the US, and dependent on Japan for critical materials. The Taylor ramp is the hedge.
3. Intel Foundry
Best for sovereign US/EU manufacturing at advanced nodes
Intel’s foundry pivot is the most interesting development in the chip foundry companies landscape. Once synonymous with manufacturing leadership, then years of delays, Intel is now shipping on its most advanced node: 18A entered high-volume production at the start of 2026.
Process Node Leadership
Intel 18A combines GAA RibbonFET and PowerVia backside power delivery, and Intel got that combination to production before TSMC and Samsung. 18A entered high-volume production at the start of 2026:
14A is the stakes-raiser. Intel committed to completing 14A development in Q2 2026, with risk production reported for 2028 and high-volume manufacturing in 2029, and has said the node’s scale depends on landing significant external customers, with the commitment window running from H2 2026 into H1 2027.
Capacity & Allocation Reality
External foundry is still new for Intel, so allocation works differently than at TSMC or Samsung:
External Customers Have Somewhat More Leverage Than At A Sold-Out Foundry…
Intel wants to win flagship AI and defense programs to validate its nodes, so it co-develops and allocates meaningful capacity against anchor wins.
Geopolitical Exposure
Intel’s structural advantage is geographic: fabs in Arizona, Oregon, New Mexico, Ireland, with Ohio under construction (now targeting around 2030 for its first New Albany fab), underwritten by the CHIPS Act and EU Chips Act. Less Taiwan and Korea dependence, more constrained on what can ship to China. For defense, aerospace, and national security buyers, this is a feature, not a bug.
4. GlobalFoundries
Best for mature-node specialty: RF, FD-SOI, automotive, Trusted Foundry
GlobalFoundries exited the 7nm race in 2018 to focus on “more-than-Moore” manufacturing: feature-rich nodes targeting applications where transistor density isn’t the constraint. That decision now looks prescient as automotive, RF, and power semiconductor demand grew faster than expected. In Q2 2026, GF posted $1.79 billion in revenue for a 3.2% share of the top-10 foundry market (TrendForce).
Process Node Leadership
GF’s portfolio runs from 180 nm down to 12 nm FinFET for select customers, with three platform pillars worth knowing:
“Leading-edge” at GF means leadership in RF, FD-SOI, photonics, GaN, and heterogeneous integration on these nodes.
TSMC vs GlobalFoundries at mature nodes: TSMC offers mature nodes at scale but allocates them behind leading-edge priorities; GlobalFoundries builds its entire business on them. For RF-SOI, FD-SOI, and US Trusted Foundry requirements, GF is the specialist choice; for plain-vanilla 28nm logic at the lowest cost, TSMC and UMC usually win on price and capacity.
Capacity & Allocation Reality
GF runs a balanced multi-region network:
Announced investments include a €1.1 billion Dresden expansion through 2028 and roughly $16 billion in US expansion across facility modernization and AI-oriented silicon photonics R&D (figures as announced in 2025; re-check GF investor relations for updates). In Q2 2026, GF posted $1.79 billion in revenue for a 3.2% share of the top-10 foundry market (TrendForce).
Geopolitical Exposure
US-headquartered with fabs in the US, Germany, and Singapore, and no fabs in China or Taiwan. GF is the only commercial high-volume foundry with US “Trusted Foundry” designation, making it uniquely attractive for secure government and defense programs that need guaranteed US-based production.
5. United Microelectronics Corporation (UMC)
Best for predictable, mature-node automotive and industrial supply
UMC is the stable middle of the foundry semiconductor market: Taiwan-based, focused on mature and mid-range nodes with strong automotive and industrial positioning, and running hot. In Q3 2026 its capacity utilization hit 92% and operating income rose 24.7% quarter on quarter, on revenue of NT$74.24 billion (US$2.32 billion).
Process Node Leadership
UMC’s specialty leadership centers on 22/28nm, used widely in automotive, industrial, networking, display, and IoT, and the 22nm business keeps setting records: 18.3% of Q3 2026 wafer sales, the highest share yet. The 14nm FinFET node delivers about 50 to 60% power efficiency improvement versus 28nm.
The Big Strategic Move Is The Intel Collaboration:
Capacity & Allocation Reality
The Customer Dynamics Differ From TSMC:
Capacity is generally available but contract-driven, with high priority for long-lifecycle automotive, industrial, and networking programs rather than speculative one-off tape-outs. Once a design qualifies on a UMC process, customers tend to stay for many years.
Geopolitical Exposure
Taiwan-based with the associated risk, but UMC runs an unusually structured supply-chain risk monitoring platform tracking critical materials, sourcing regions, and geopolitical events. The Intel 12nm partnership adds Western capacity for select products. A magnitude 7.4 quake in eastern Taiwan in 2024 temporarily halted production at both TSMC and UMC fabs, a reminder of the structural exposure.
6. Semiconductor Manufacturing International Corporation (SMIC)
Best for China-centric designs requiring domestic fabrication
SMIC is China’s flagship foundry and the center of Beijing’s semiconductor self-sufficiency push. It is strong and expanding at mature nodes, pushing constrained 7nm-class capacity without EUV, and increasingly a domestic story: China accounted for 90.2% of SMIC’s revenue in Q2 2026, up from 84.1% a year earlier.
Process Node Leadership
SMIC offers foundry services from 350nm down to 7nm-class technologies, with most commercial volume at 28nm and above.
The N+1 process:
Advanced-node capacity is tiny relative to demand and heavily directed:
SMIC became a key foundry for Huawei and other sanctioned Chinese firms after TSMC stopped taking their orders.
Advanced output is directed toward domestic priorities, with Huawei the most prominent beneficiary.
Capacity & Allocation Reality
Q2 2026 was SMIC’s first quarter above $3 billion:
AI-adjacent demand (power management, interface, and supporting chips around AI systems) is what’s filling the fabs.
Geopolitical Exposure
SMIC was added to the US Entity List in 2020 with expanded controls since. Export restrictions limit access to advanced equipment (notably EUV), complicate servicing for existing tools, and stretch node migration timelines. For US, EU, or Japan-based fabless companies, SMIC carries elevated regulatory and reputational risk. For China-centric buyers, it is increasingly the only realistic local option.
7. Texas Instruments
Best for long-lifecycle analog and embedded for industrial and automotive
TI isn’t a foundry in the TSMC sense. It is a large analog and embedded IDM running 45 to 130nm-class fabs at scale, primarily for its own products. Worth understanding the distinction before evaluating.
Process Node Leadership
TI exited digital logic process development at 45nm, choosing to stop internal node scaling and rely on external pure-plays for 32nm and below where needed. Today TI focuses its internal manufacturing on analog and embedded products using 45 to 130nm processes on 300mm wafers, explicitly optimized for analog and embedded rather than chasing 7nm or 5nm.
Capacity & Allocation Reality
TI is midway through roughly $60 billion in US-centric expansion across seven fabs at three sites: Sherman and Richardson in Texas plus Lehi in Utah, all focused on analog and embedded 300mm wafers.
Geopolitical Exposure
8. Infineon Technologies
Best for automotive power, SiC, GaN, and security MCUs
Infineon is a vertically integrated power, automotive, and security IDM with its own fabs and specialty processes, so it is not a general-purpose logic foundry. Worth knowing what you’re buying: product supplier with selective manufacturing partnerships, not open foundry access.
Process Node Leadership
Core strengths sit in:
On CMOS logic, Infineon uses 40nm and 28nm nodes for security controllers and automotive/IoT MCUs.
The Wide-Bandgap Leadership Is The Real Differentiator:
Infineon was the first semiconductor manufacturer to develop 300mm GaN power wafer technology on existing high-volume infrastructure, significantly increasing potential output per wafer. SiC capacity gets first call on capital allocation alongside strategic fabs and R&D.
Capacity & Allocation Reality
Manufacturing is centered in Europe (Germany and Austria), with strategic 300mm power and wide-bandgap investments there. Capital allocation prioritizes strategic fabs, SiC capacity, and R&D for automotive, industrial power, and renewables.
Geopolitical Exposure
9. NXP Semiconductors
Best for automotive platforms spanning mature to advanced nodes
NXP is a fab-light hybrid IDM: it designs mostly automotive and embedded chips and uses a mix of in-house mature fabs and outsourced advanced nodes (TSMC, GlobalFoundries, joint ventures). Not an open foundry, but a platform supplier with smart manufacturing partnerships.
Process Node Leadership
NXP uses in-house fabs for analog and mixed-signal and outsources advanced digital logic to foundry partners. The roadmap talk became product in 2026: at CES 2026, NXP unveiled the S32N7 super-integration processor series on a 5nm foundation, following the 5nm S32N55, putting 5nm automotive silicon in the market, with 3nm automotive parts on the roadmap through foundry partners.
Capacity & Allocation Reality
You can’t book NXP fab slots as a fabless customer. You become an NXP chip customer or, rarely, a JV or technology partner.
Geopolitical Exposure
10. STMicroelectronics
Best for European automotive, industrial, and FD-SOI MCUs
ST is an IDM focused on automotive, industrial, power, and MCUs, with European fabs. Like NXP and Infineon, it is a product-centric manufacturer, not an open leading-edge logic foundry.
Process Node Leadership
ST runs a wide range of mature and mid-range CMOS nodes (130/90/65/55/40/28nm) with digital, mixed-signal, and RF options for analog, power, and microcontroller products.
The Headline Development:
The STM32V8, launched in November 2025, is the industry’s first 18nm microcontroller: an Arm Cortex-M85 at 800MHz with 4MB of embedded phase-change memory (ePCM), built on ST’s 18nm FD-SOI process at its 300mm Crolles fab in France, with broader availability from Q1 2026. SpaceX selected it for the Starlink constellation.
The 18nm FD-SOI + ePCM process delivers around 50% better performance-per-power than ST’s 40nm embedded NVM node and 2.5x higher NVM density. ST states it is the only sub-20nm process supporting 3V operation, enabling integrated analog (power management, reset, clock, ADC) for robust industrial and automotive environments.
Capacity & Allocation Reality
Historically during capacity crunches, ST has prioritized strategic partners and top customers.
Geopolitical Exposure
EU-headquartered (Geneva, with major operations in France and Italy) with a diversified supply chain across Europe, Taiwan, Japan, and the US.
Comparison: Top 10 Semiconductor Foundries 2026
Legend:
✔️ Full
⚠️ Partial / limited
❌ Not supported
Semiconductor Foundry Market Share, Q2 2026
The top 10 semiconductor foundries earned a record $53.49 billion in Q2 2026, up 11.5% quarter on quarter, per TrendForce’s September 2026 ranking. The five largest by revenue share:
Source: TrendForce, September 9, 2026. The notable movement: SMIC grew 20% in the quarter and is closing on Samsung for second place, while TSMC’s share set another record.
Watchlist for 2027
Two names sit outside this top 10 but belong on a 2027 shortlist.
Rapidus is Japan’s state-backed 2nm bet. Its IIM-1 pilot line in Chitose, Hokkaido has run since April 2025, a 2nm gate-all-around prototype built on IBM’s nanosheet architecture reached its expected electrical characteristics in July 2025, and the process design kit reached early customers in Q1 2026. Mass production is targeted for 2027, the company is reportedly in talks with more than 60 potential customers with IBM, Tenstorrent, and Broadcom named as leading prospects, and 1.4nm development starts in 2026. If the 2027 ramp lands, it becomes the first new leading-edge entrant in decades.
Vanguard International Semiconductor (VIS), a Taiwan-based mature-node specialist, re-entered eighth place in TrendForce’s Q2 2026 ranking on power management IC demand, a reminder that the mature-node tier below this list keeps consolidating around a few reliable names.
Which Foundries Have the Best Yield Performance in 2026?
No leading foundry publishes official yield data, so the honest comparison is reported maturity:
The practical read for buyers: TSMC for lowest yield risk today, Samsung and Intel for leverage and headroom, with the gap narrowing through 2027.
Leading European Foundries for Mature-Node Mixed-Signal
European mature-node capacity centers on four names:
For hyperscaler-scale AI accelerators, the 2026 allocation math is simple: TSMC N2/N3 capacity is committed to anchor customers first, so new large programs either pay for priority, split volume across TSMC and Samsung (the Tesla AI5 model), or anchor Intel’s 18A-P/14A ramp in exchange for influence over it. Dual-foundry designs cost more up front and remove the single point of failure that a sold-out node represents.
How Do Semiconductor Foundries Compare on Sustainability?
Foundries do not publish comparable third-party sustainability ratings, so the practical comparison is public commitments and annual ESG reports. UMC, for example, states a net-zero emissions target by 2050 and runs supplier decarbonization programs with partners including Infineon, per its 2026 Q1 reporting. Buyers with Scope 3 reporting obligations should request each foundry’s latest ESG report during qualification rather than relying on rating aggregators.
Where AI Visual Inspection Fits In Semiconductor Manufacturing
Whichever foundry partner you choose, defect detection at submicron scale determines yield economics, and yield is what separates the foundries above. Averroes ai plugs into existing inspection equipment, feeding AI defect classification and segmentation directly into your process control loop, no new hardware required. Teams tracking fab output alongside partner selection can pair this page with our guide to manufacturing yield optimization.
The production evidence: in photomask manufacturing, Averroes sustained 99.993% classification accuracy across 1.4 million production images over 13 months at Tekscend Photomask, with defect review running 8x faster than the manual baseline.
Ready To Stop Losing Wafers To Escapes?
99%+ defect detection accuracy with near-zero false positives.
Semiconductor Foundries FAQs
What is the difference between a foundry and an IDM?
The difference between a foundry and an IDM comes down to who designs the chips. A pure-play foundry like TSMC fabricates chips designed by other companies (fabless customers). An IDM like Intel, TI, or Infineon designs and manufactures its own chips. Intel and Samsung run hybrid models where they do both.
Which semiconductor foundries can manufacture 2nm chips?
The semiconductor foundries in volume production at 2nm in 2026 are TSMC (N2, 8% of its wafer revenue by Q3 2026) and Samsung (SF2, with trial 2nm production also running at its Taylor, Texas fab). Intel’s 18A is a 2nm-class node with GAA and backside power delivery in high-volume production, offered externally through its 18A-P variant. Japan’s Rapidus targets 2nm mass production in 2027.
Why is TSMC so dominant in the semiconductor foundry market?
TSMC’s dominance in the semiconductor foundry market comes from three compounding advantages: process leadership at leading-edge nodes, a deep ecosystem of EDA tools and IP partners, and trust built over decades with fabless customers like Apple, Nvidia, and AMD. As of Q2 2026 it holds a record 72.5% of top-10 foundry revenue and roughly 90% of leading-edge production.
Who are TSMC’s main competitors?
TSMC’s main competitors are Samsung Foundry (5.9% share in Q2 2026, first to 2nm GAA mass production), Intel Foundry (18A in high-volume production, the sovereign US option), and SMIC (5.4% share, China’s domestic champion). At mature nodes, UMC and GlobalFoundries compete on specialty processes. Japan’s Rapidus becomes a leading-edge challenger if its 2027 2nm ramp lands.
Which foundry has the best yield in 2026?
No foundry publishes official yield data, but reported figures put TSMC ahead: initial 2nm yields of 70 to 90% at volume start in late 2025, versus Samsung’s 2nm reported crossing the roughly 60% viability threshold in 2026 and Intel’s 18A reported reaching industry-standard yields by early 2027. Mature-node yields at UMC and GlobalFoundries are stable and rarely the deciding factor.
How long does it take to qualify a new semiconductor foundry partner?
Qualifying a new semiconductor foundry partner typically takes 12 to 18 months for a new node tape-out and significantly longer for automotive-grade qualification. Advanced-node capacity also has to be reserved well ahead: as of 2026, leading-edge allocations are reportedly booked a year or more out, with prepayment common for strategic programs.
Conclusion
Ten semiconductor foundries, ten different bets on what matters most in 2026. TSMC owns leading-edge AI and HPC but charges scarcity premiums and concentrates Taiwan risk. Samsung delivers the credible #2 option with tight memory integration and US 2nm capacity now running trial wafers in Texas. Intel Foundry is the sovereign manufacturing play that finally has a high-volume node to back it. GlobalFoundries and UMC anchor the mature-node and specialty market for buyers who don’t need 3nm to win. SMIC serves China-centric designs at high regulatory cost. TI, Infineon, NXP, and STMicroelectronics are IDMs to buy from, not tape out at.
Whichever foundry partner you pick, yield economics decide whether your design ships on margin or burns through reserves. Averroes catches 40 to 60% more submicron defects on your existing inspection equipment. Book a free demo to see how AI visual inspection changes your foundry yield math.