Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Both statements are true: TSMC has demonstrated electrically working complementary field-effect transistors (CFETs), yet the company has not announced a CFET production node or launch date. “Working” describes a functioning research device; it does not mean a repeatable, economical process that can manufacture billions of transistors with acceptable yield.
TSMC made the original disclosure at its 2023 European Technology Symposium, describing CFETs as several generations from mass production. Later coverage reported a more advanced milestone—a working CFET inverter—but still identified major integration, alignment and yield obstacles. The latest public material cited here does not establish a TSMC CFET high-volume-manufacturing date.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Introduction to Semiconductor Manufacturing Technology, Second Edition | $129.00 | Buy on Amazon |
| 2 |
|
Chip War: The Fight for the World's Most Critical Technology | $15.75 | Buy on Amazon |
| 3 |
|
Semiconductor Devices: Theory and Application | $17.52 | Buy on Amazon |
| 4 |
|
Semiconductor Manufacturing Technology | $146.64 | Buy on Amazon |
As an Amazon Associate I earn from qualifying purchases.
What TSMC actually announced
In comments from TSMC’s 2023 European Technology Symposium, reported by AnandTech, the company said it had working CFETs in its laboratories. TSMC did not attach the demonstration to a named commercial node, give a production year or promise that CFET would be its next transistor architecture.
The same report said nanosheet gate-all-around (GAA) transistors were expected to serve for multiple generations before CFET became relevant. “Generations away” was a broad roadmap description, not a fixed count or calendar date. TSMC was describing ongoing research into options beyond nanosheets, not announcing a locked-in manufacturing plan.
What CFET means
CFET stands for complementary field-effect transistor. Modern CMOS logic combines two complementary transistor types:
- nFET/NMOS: primarily uses electrons as charge carriers.
- pFET/PMOS: primarily uses holes as charge carriers.
- CMOS logic: uses the two networks so one pulls a signal high while the other pulls it low.
In planar CMOS and FinFET designs, the nFET and pFET generally occupy adjacent positions on the silicon surface. A CFET places the complementary devices vertically, allowing them to share more of the same horizontal footprint. It is therefore mainly a three-dimensional integration architecture—not simply a smaller version of an ordinary transistor.
CFET and GAA are related, not interchangeable
In a GAA nanosheet transistor, the gate surrounds each channel on all sides, improving electrostatic control compared with a FinFET. A CFET changes how the n-type and p-type devices are arranged: one is placed above the other. A CFET could use nanosheet-like channels and GAA gates, so it is more accurate to view CFET as a possible next-level arrangement built on advanced transistor structures than as a direct replacement for GAA.
Free tools Windows power users keep installed
One-click scans. No signup required.
Why the industry is interested
Vertical complementary stacking could offer several benefits, but each remains conditional on real process results.
- Higher logic density: putting nFET and pFET devices in the same footprint may shrink standard cells.
- Potentially shorter local wiring: vertical proximity can reduce some lateral connections.
- Possible power and performance gains: less wiring may reduce certain resistance and capacitance, although contacts, parasitics and thermal effects can offset those gains.
- A route for continued scaling: CFET is one candidate for extending logic scaling after multiple generations of nanosheet GAA technology.
These are projected advantages, not measured specifications for a shipping TSMC product. The outcome depends on the complete device, interconnect and circuit design. Industry roadmap analysis places CFET after nanosheet and possibly forksheet architectures, while also emphasizing that foundries may choose different sequences (Semiconductor Engineering).
Why a working lab device is not a production process
A laboratory test can use a small number of carefully selected structures, specialized process steps or unusually intensive tuning. Commercial manufacturing must make the same structures repeatedly across entire wafers, many wafers and many lots, while meeting electrical, reliability and cost targets.
The manufacturing hurdles
- Wafer-scale repeatability: selected working devices are not evidence that every die works.
- Yield: each additional deposition, etch, release, alignment and contact step creates new defect opportunities. A process must produce enough good dies to justify its wafer cost.
- Overlay and alignment: the upper and lower transistors must line up within very tight tolerances. Misalignment can alter channels, gates, contacts and local wiring.
- Contacts and resistance: the devices need low-resistance connections to source, drain, gates, power rails and signal wiring. A stacked structure is useful only if those connections do not consume its area or performance advantage.
- High-aspect-ratio processing: reaching and connecting features in a three-dimensional stack makes etching, filling and inspection more difficult.
- Thermal behavior: the upper and lower devices may experience different temperatures, and heat generated in one can affect the other.
- Reliability: products must survive voltage, temperature, electromigration, bias-temperature-instability and lifetime testing.
- Design enablement: customers require standard-cell libraries, SPICE models, design rules, parasitic extraction, process-design kits and signoff flows.
A device that turns on is therefore an early physics milestone. Industrialization requires a complete process and ecosystem.
Why contacts and local routing are especially difficult
CFET integration is not solved by placing one transistor on top of another. The process must connect the source and drain regions of both devices, control both gates, distribute power and route logic signals without excessive resistance, capacitance or area overhead.
Coverage of TSMC’s IEDM 2024 work reported a working CFET inverter and a method for forming a local interconnect between the upper and lower devices. That is a more meaningful circuit milestone than an isolated transistor, but the same analysis identified tight alignment and high-aspect-ratio structures as continuing high-volume-yield challenges (SemiAnalysis).
What the inverter demonstration proves—and does not prove
An inverter contains complementary pull-up and pull-down devices and produces a real logic function. Demonstrating one with stacked devices indicates that the nFET and pFET can be integrated sufficiently to operate together, including a local connection between them.
Rank #3
It still does not establish a qualified logic process. The public report does not provide a TSMC production date, wafer-scale yield, long-term reliability data, customer PDK or the economics of manufacturing complete chips with that flow.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11What likely comes before CFET
- FinFET: the established three-dimensional channel architecture used before gate-all-around devices.
- GAA nanosheet: gates surround horizontal channels for stronger electrostatic control.
- Possible forksheet stages: variants that reduce the separation between nFET and pFET without fully stacking them.
- CFET: vertical complementary stacking, if alignment, contacts, thermal behavior, yield and cost become manageable.
- Further alternatives: new channel materials and other three-dimensional devices may compete with or complement CFET.
This is an industry-level sequence, not a guaranteed TSMC roadmap. TSMC’s own 2023 comments specifically placed multiple nanosheet generations before CFET adoption became relevant (AnandTech).
Lithography is only one part of the problem
The original reporting identified extremely precise lithography, potentially including High-NA EUV, among the challenges for future CFET integration. High-NA tools may help pattern critical features, but no evidence establishes that every CFET implementation requires them. The full flow also depends on deposition, etch, alignment, contacts, interconnects, metrology and inspection.
High-NA EUV brings its own engineering and economic considerations, including field-size and reticle constraints. Coverage of future TSMC nodes has treated its adoption, and the timing of CFET, as uncertain rather than settled (Tom’s Hardware).
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Why a node number cannot predict CFET timing
Labels such as “2 nm,” “1.4 nm” or “angstrom” are not direct measurements of gate length and do not certify a particular transistor architecture. Foundries can introduce new devices, backside power delivery, buried rails, interconnect changes, packaging improvements and new standard-cell layouts on different schedules.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →CFET is one element of a broader scaling strategy that also includes backside power, advanced packaging, chiplets, memory integration and system-level co-optimization. A named node therefore cannot, by itself, prove that CFET has been adopted.
How to judge future CFET announcements
| Readiness question | What stronger evidence would show |
|---|---|
| Device demonstration | An individual CFET operates electrically. |
| Circuit demonstration | An inverter or larger logic block functions with complementary stacked devices. |
| Integration | Top and bottom devices are connected with a repeatable, manufacturable flow. |
| Performance | Drive current, leakage, speed and power are reported under defined conditions. |
| Variability and reliability | Across-wafer distributions and lifetime testing are disclosed. |
| Yield and economics | Manufacturing yield, wafer cost and defect data support volume production. |
| Design ecosystem | Libraries, models, rules, PDKs and EDA verification flows are available. |
| Customer qualification | Customers design and validate products on the process. |
CFET is not the only scaling option
Foundries can pursue additional nanosheet generations, forksheet devices, backside power delivery, buried rails, advanced interconnects, two-dimensional channels, chiplets, 2.5D/3D packaging, memory integration and specialized accelerators. These approaches address different bottlenecks; packaging or power delivery can improve a product before a new transistor architecture reaches high-volume manufacturing.
What the announcement means as of August 2026
The established facts are narrow but important: TSMC reported working laboratory CFETs in 2023, later coverage described a working CFET inverter, and the company has not publicly committed in the cited material to a production node or date. The projections concern potential density, wiring, power and performance benefits. Speculation begins when those results are converted into a specific year, node number or guaranteed product advantage.
The next convincing milestones would be larger logic blocks, wafer-scale demonstrations, published variability and reliability data, quantified contact and interconnect resistance, evidence of usable yield, early design collateral and an explicit TSMC manufacturing commitment.
Do these 3 things before closing this tab:
1Fix the driver behind crashes, sound loss and screen glitches2Clear out junk files and repair common Windows errors3Scan for outdated or missing drivers - takes under a minuteThe Bottom Line
TSMC’s result shows that CFETs are technically viable enough to justify serious industrial research. It does not show that they are ready for mass production. The decisive breakthrough will be a repeatable, affordable process that can connect and manufacture complete logic at wafer scale—not merely another transistor that turns on.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
