Pleasing progress at TSMC's Arizona location has resulted in company leadership expediting a "push into advanced node deployment." Mid-October reports cited an increased domestic demand for 2 nm-class wafer products. Despite the foundry industry leader and NVIDIA's public celebration of debut "made in the USA" NVIDIA "Blackwell" wafers, TSMC's Phoenix-adjacent campus lacks in one key area: advanced packaging. Not long after the conclusion of festivities, computer hardware news outlets highlighted Fab 21-produced (4 nm) Blackwell products being sent overseas—specifically, back to Taiwan—for Chip on Wafer on Substrate (CoWoS) treatments.

Later on in October, TechPowerUp's AleksandarK covered the importance of advanced packaging on North American soil. His special editorial piece focuses on Intel Foundry's potential advantages in this area of expertise; thus poised to take a leading position within the region's semiconductor industry. Yesterday, Taiwan's Liberty Times outlined a major near-future change for one section's of TSMC Arizona's vast spread: "(leadership) plans to convert the land of the planned sixth phase fabrication plant (P6 fab) into an advanced packaging (AP) plant. If the construction goes smoothly, it is expected to be installed as early as the end of 2027 and enter the pre-production stage." Some of the claimed inside track timelines seems to roughly align with previous reports of TSMC and Amkor Technology's $7 billion forthcoming "advanced outsourced semiconductor packaging and testing" facility.

Last month, ASUS started a teaser campaign that alluded to a brand-new ROG Mini motherboard design. Promo material provided a very pixellated preview of a pale Mini-ITX form factor product. Earlier this week, a single render leaked out—image quality was good enough to identify the unreleased SKU as a white spin-off of the already launched Strix B850-I Gaming WiFi7 board design. At some point within the past couple of days, the main ASUS web presence was updated with a dedicated "W" model product page. We suspect that their next move will be the publication of press material, unless staffers deem this SKU's slight redecoration unworthy of such attention. Annoyingly, the White Edition's detailed web portal does not provide a clear view of the board's back section.

Unsurprisingly, the Taiwanese manufacturer has rolled out the standard black variant's exact same description and specs: "(our latest model) integrates cutting-edge technology into a sleek Mini-ITX form factor. It features PCIe 5.0 slots for graphics and storage, dual-channel DDR5 slots and robust power delivery for top-tier gaming and heavy workloads. Meanwhile, advanced cooling solutions maintain peak performance, and Wi-Fi 7 ensures ultra-fast connectivity. This platform delivers the power and connectivity that advanced AI PC applications demand." One major difference, at least on digital paper, is the White Edition's deployment of a "64 MB BIOS." The larger ASUS BIOS chip is advertised as offering a "more powerful interface" and "support for future AM5 processors."

ASRock's Vietnamese website has added an intriguing new microATX motherboard to its portfolio of Intel H610 chipset products. Uniko's Hardware stumbled upon this low-key addition earlier today—their social media post expressed excitement about the young SKU's unique RAM setup. They proclaimed: "a new LGA 1700 H610M motherboard that supports DDR4 and DDR5—but not at the same time! Six DIMM slots on this mobo = two for DDR4, and the other four support DDR5, both in dual channel mode...ASRock saving our ass!" In-depth investigations reveal that the H610M COMBO's first BIOS release is tagged with a June 2025 publication date.

This basic board design's hybrid slot system—targeting office and industrial applications—offers refreshing flexibility. Over the past couple of weeks, computer memory prices have escalated. In particular, mainstream DDR5 product costs of ownership have hit unreasonable heights. Fortunately, the H610M COMBO can accept cheaper DDR4 sticks. Uniko's Hardware recalls that an obscure ONDA H610M+ board offered a similar system three years ago, albeit restricted to one slot per DDRx type (single channel). It is encouraging to see a mainstream manufacturer release a more capable "combo" design, but the new model's specifications and feature set outline entry-level operations—no pricing information or retail availability has emerged online. Hardcore gamers and overclockers will likely sniff at the H610M COMBO SKU's basic capabilities.

Artificial intelligence works fast, but its energy consumption is growing rapidly. A German-Taiwanese research team is now developing a solution: new memory for leading chip technologies smaller than 3 nm. These innovative nanosheet devices enable computing operations directly in memory, thereby drastically reducing energy consumption. They are based on ferroelectric field-effect transistors (FeMFETs) made from hafnium oxide, which are particularly efficient. With a joint research program, Fraunhofer IPMS, Fraunhofer IMWS, and the Taiwanese research institute TSRI are laying the foundation for the next generation of energy-efficient AI chips - from smartphones and automobiles to medical devices.

Given the rapidly growing demand for artificial intelligence (AI) and neuromorphic computing, the energy consumption of data centers and edge systems is increasing dramatically. A key bottleneck is the transfer of data between main memory and the computing unit. A joint German-Taiwanese project aims to address precisely this issue: innovative memory technology will enable computing "directly in memory", with significantly lower latency and energy consumption.

Earlier today, an intriguing 30-second-long video was uploaded to the Chinese Bilibili platform. A brief sequence seems to demonstrate a refreshed Gigabyte AORUS BIOS user interface, with familiar AMD technologies visible across multiple windows. The leaker is an amusingly named individual: "Coward Yin Zhiping." Their blog/channel shows a history of BIOS tweaking and overclocking across several brands. Fittingly, humor is injected into video descriptions—the latest upload is titled with deflection in mind: "ASUS hired aliens as designers; their new BIOS design is simply deadly." Hours later, Uniko's Hardware published their interpretation of Zhiping's leak. Screenshots show a completely new style of presentation, different from current iterations and Gigabyte's September previews—according to VideoCardz.

Diagnostic data indicates a platform based on Team Red's 800 chipset series. Specifically, an X870E AORUS MASTER X3D ICE board paired with a Ryzen 9 9900X processor, and 48 GB (2 x 24 GB) DDR5 (Hynix) memory. The supposed revised UEFI has "X3D 2.0" and "PBO" tuning features categorized under "Black technology." A new "Cookie score" seems to be a completely new addition—Uniko's Hardware believes that it is an: "(AI-powered) CPU quality assessment design by HiCookie, a legendary overclocker and engineer at GIGABYTE." Several discrepancies were highlighted by initial investigators, perhaps pointing to unfinished elements. The video and screengrabs show a months old session date: September 23, 2025.

Acer Inc., Esquarre Vision Limited, and OPTOELECTRONICS CO., LTD., a Japanese AIDC (Automatic Identification and Data Capture) solution provider known as Opticon, jointly announced plans for Acer and Esquarre to invest approximately 38% and 20%, respectively, in Opticon for JPY 1,589,419,104 and JPY 819,000,000. The transaction is expected to be completed in early 2026, supporting Opticon's growth and global expansion.

For Acer, the move expands its portfolio, adding identification and machine vision solutions that will create synergy with its subsidiaries AOPEN (TWSE: 3046) and Posiflex (TWSE: 8114). AOPEN is a company which focuses on industrial computers for display engines and edge computing, while Posiflex Group is a global leader in Commercial AIoT, delivering smart online-to-offline (O2O) integrated solutions and scenario-defined embedded edge computing platforms.

ASUS Republic of Gamers has teased the imminent arrival of a spun-off STRIX B850-I GAMING WIFI motherboard model. The standard black version has been out for almost a year, but a pale sibling will be added to the Taiwanese manufacturer's collection of small form factor (SFF) mainboards. A cute mascot pixel art-themed teaser was released prior to the discovery of a single product promo render. In the early hours, realVictor_M uploaded this lone image to social media. As expected, the forthcoming white Mini ITX product—dubbed "ROG STRIX B850-I Gaming WIFI7 W"—deploys a familiar layout.

The ASUS mainboard design team has revised a small portion of visual elements; mainly on the I/O and M/2 covers/shrouds. Branding and graphical themes are slightly different, when compared to material printed on the standard black SKU. We expect similar visual updates to be present on the white edition's rear surface, but this morning's leak did not include any additional shots of the motherboard's back and sides. At the time of writing, official ASUS web presences do not list this SKU. It is possible that the leaked render was sourced from an inadvertent webstore entry.

Giga Computing, a subsidiary of GIGABYTE Group and a leader in high-performance computing and data center solutions, today announced a strategic partnership with Syrma SGS Technology Limited, a premier Indian electronics manufacturing services provider. This collaboration will establish local manufacturing of GIGABYTE server products in Tamil Nadu, India, supporting the country's "Make in India" initiative and strengthening Giga Computing's supply chain presence in South Asia.

To enhance global advanced manufacturing capabilities and create a more agile supply chain, Syrma SGS will begin assembly of GIGABYTE server motherboards at its Chennai facility to meet India's growing market demand. Initial production will include enterprise-grade boards such as the GIGABYTE MS73-HB0 and MZ33-AR1, delivering critical hardware for data centers and enterprise computing. The partnership will later expand to full box-build system manufacturing with export capabilities, enabling Giga Computing to accelerate global delivery with greater efficiency and flexibility.

GEEKOM has officially announced the upcoming launch of its GeekBook X series, scheduled for mid-December 2025. Designed to balance portability and performance, the new lineup aims to elevate productivity, creative workflows, and everyday entertainment. Leading the charge is the GeekBook X14 Pro, which sets a new benchmark as the lightest Windows laptop in its class, redefining standards for mobile computing and professional efficiency.

Founded in Taiwan in 2003, GEEKOM has emerged as a formidable player in the premium mini PC market. In Q3 2025, the company recorded a remarkable 150% revenue growth over the previous quarter. Now expanding into the notebook segment, GEEKOM is positioning itself as a full-spectrum PC brand, encapsulated in its brand vision: "Think PC, Think GEEKOM." The company now competes alongside global tech giants -yet in many European countries, GEEKOM's high-end mini PCs consistently outperform these rivals in popularity.

Leaked marketing material indicates that ASRock is readying a new branch of Mini PCs, that will exist alongside already launched DeskMeet, DeskMini, and Jupiter ranges. Yesterday afternoon, a specification sheet was shared online—by well known hardware sleuth momomo_us. The Taiwanese manufacturer seems to have an Intel-flavored DeskSlim B760 model prepared for imminent release. This system is centered on Team Blue's familiar LGA 1700 platform, covering 12 to 14th generations of Core processors. Fittingly, an Intel B760 chipset controls proceedings. Alternatively, customers can choose X600 variants—AMD hardware is housed in the same DeskSlim enclosure; a compact sub-5 liter prospect.

According to the leaked ASRock specification sheet, their Team Red variant—based on an X600 chipset, utilizing the AM5 platform—can accommodate Ryzen 7000, 8000 and 9000 series CPUs. Early analysis posits that the Intel and AMD systems will be sold as barebones products. Users can add their own appropriately-sized graphics cards—maximum would be a dual-slot low profile design, 200 mm in length. A standard 240 W external power adaptor is listed as a bundled-in item, but the spec sheet mentions a more potent 330 W optional unit. VideoCardz believes that ASRock will reveal official details next week; with brand new DeskSlim products actively listed across web presences.

ASUS has announced that, in collaboration with the National Center for High-performance Computing (NCHC) to build the new AI supercomputer, the facility is now officially in operation. It features a dual-compute architecture that includes the cluster and the latest NVIDIA GB200 NVL72 system—Taiwan's first fully liquid-cooled AI supercomputer deployment of this architecture. The Nano4 NVIDIA HGX H200 system delivers up to 81.55 PFLOPS of performance and is ranked No. 29 on the TOP500 list, marking a meaningful milestone in strengthening Taiwan's AI-computing capabilities and accelerating its intelligent transformation.

NCHC spokesperson Dr. Chia-Lee Yang said: "As the designer of the Nano4, NCHC drove the architecture, liquid-cooling strategy, and system integration. Along with AI-infrastructure expertise from ASUS, this newly built system feeds advanced computes for generative AI, big data, deep learning, and HPC, unlocking unprecedented computational capabilities for both academia and industry."

Earlier this week, Tim Culpan (of Culpium newsletter fame), claimed that TSMC Arizona's flagship Fab 21 production facility experienced a major power cut around mid-September. The Taipei-based journalist has gathered information about this unfortunate incident from unnamed semiconductor industry figures. An unwanted interruption—supposedly caused by an outsourced industrial gas vendor—affected: "the flow of crucial inputs needed for chip making, forcing the facility to shut down for at least a few hours...As a result, the company had to scrap thousands of wafers that were in production for clients at the site which include Apple, NVIDIA, and AMD." Earlier today, a TSMC representative confirmed—to Tom's Hardware—that a power outage did occur, but did not disclose any details regarding the loss of wafers.

The Culpium insider article loosely connects the aforementioned incident to an apparent drop in TSMC Arizona profits—around the third quarter of this year. Culpan outlined some minor fallout effects: "impact to clients is likely to be negligible, I was told, and the financial loss to TSMC may be covered by insurance. Capacity at Fab 21 is still quite small, and many products being made there have already been taped out and manufactured in Taiwan previously. In past disruptions, lost production and revenue was made up in the subsequent quarter." Roughly ten months ago, the semiconductor giant's home operations were affected by an earthquake. Industry moles reckoned that (up to) 20,000 wafers were damaged during this wave of seismic activity; deemed to be very minor setback. The various fabs apparently bounced back swiftly into 24/7 production mode.

Foxconn's Wisconsin manufacturing hub has received a thumbs up from local government officials; after considering a proposed $569 million investment. Major additional funding will be directed into the expansion of the Taiwanese's giant's Racine County operation. According to a fresh Reuters news article, company officials cited a growing demand for AI server equipment—the facility's upcoming upgrades are deemed necessary to strengthen supply chains in the USA. Foxconn's relationship with the Wisconsin Economic Development Corporation (WEDC) seems to be in a better place nowadays; the two organizations have reportedly tussled over tax credits in the recent past. Within that time frame, the multinational electronics manufacturer has claimed that it has spent roughly $2 billion on upkeep of its Mount Pleasant-based site.

Jerry Hsiao, Foxconn's chief product officer, stated to Reuters (today): "as the demand for more data infrastructure continues to rise, Foxconn will keep responding to our customers' needs with flexibility and at scale in the United States." In addition, the boss estimates that the Wisconsin facility's team accounts for about a fourth of the company's total North American workforce. Looking to the future, he believes that the latest cash injection will spur on the creation of 1300+ new roles—by 2030. Original plans for the mega factory—going back to the mid-2010s—focused on the production of LCD panels, but present day market trends have caused a pivot to the pumping out of AI-oriented datacenter gear.

Leadtek is best known for its professional-grade card manufacturing partnership with NVIDIA, but the Taiwanese firm surprised many observers with its low-key Q1 2025 reveal of modern gaming products. Despite being a late addition to the "Blackwell" generation party, the veteran graphics card producer debuted a comprehensive range of new HURRICANE WinFast designs. At the time, they debuted RTX 5090, 5080, and RTX 5070 Ti models. This week, VideoCardz happened upon a freshly-published WinFast RTX 5060 HURRICANE product page. This non-overclocked entry-level card seems to be present within TechPowerUp's GPU database, although it has clearly missed a claimed "May 19th" launch date.

Refreshingly, Leadtek has opted to decorate their lowly dual-fan/dual-slot 8 GB offering with fetching white livery—distinguishing it from the rest of the company's gray/black HURRICANE pack. VideoCardz reckons that Leadtek has concentrated mainly on the Asian market, hence a complete lack of current-gen cards across Western territories. Oddly, there is little evidence of RTX 50-series HURRICANE ownership being shared online. Since March, Leadtek has expanded the range with RTX 5070 and 5060 Ti options. Various official web entries do not allude to pricing or regional availability.

Liberty Times reports that TSMC is preparing to build three more 2 nm fabs in Taiwan, with a combined investment of roughly NT$900 billion (about $28 billion). The sites will likely be located near the Southern Taiwan Science Park. According to the report, TSMC originally planned seven 2 nm fabs, two in Hsinchu's Baoshan district and five in Kaohsiung's Nanzi district. The three new facilities would raise the total to 10, as demand for advanced nodes continues to climb. Groundbreaking could start as early as next year, pending land acquisition and environmental approvals. TSMC began initial 2 nm production this quarter and is already expanding its next nodes. The company broke ground on its first 1.4 nm fab in Taichung in November with volume production aimed for the second half of 2028. The A14 (1.4 nm) node is TSMC's next major process milestone after the 2 nm node and is expected to deliver up to 15% higher performance at the same power. Alternatively, at similar speeds to the 2 nm process, the 1.4 nm node will use 30% less power.

The company is also accelerating its U.S. plans. During its October earnings call, Chairman C.C. Wei said TSMC will bring N2 and more advanced technology to Arizona faster than originally scheduled. Nikkei recently reported that TSMC is close to securing more land next to its latest Arizona fab, signaling further expansion beyond its current $65 billion investment. As previously reported, TSMC expects the third U.S. fab to start 2 nm and A16 production in 2027, one year earlier than planned. The TSMC push on its U.S. expansion strategy is likely a result of competition from Intel's 18A and the upcoming 14A, however, U.S. fabs will remain one node generation behind the Taiwanese ones.

During Foxconn's annual tech day in Taipei, the company announced that its $1.4 billion supercomputing center, built in partnership with NVIDIA, will be ready in the first half of 2026. The facility will become Taiwan's largest advanced GPU cluster, powered by NVIDIA Blackwell GB300 hardware and operated under Foxconn's new AI and cloud unit, Visonbay.ai. NVIDIA VP Alexis Bjorlin said that with GPU performance rising so quickly, many companies will find it more economical to rent compute capacity instead of building their own facilities. Foxconn plans to invest $2-3 billion per year in AI as Chairman Young Liu says the company can already build around 1,000 AI racks per week, a number set to rise next year.

In a separate announcement, OpenAI confirmed a new partnership with Foxconn to co-develop hardware for data centers. Foxconn will design and manufacture components such as cabling, power systems, and other infrastructure gear. While the agreement doesn't include purchase commitments, OpenAI will get early access to test the hardware. Foxconn's deepening collaboration with OpenAI mirrors the model it already uses with Google, AWS, and Microsoft. The partnership gives Foxconn better feedback on rack design and helps speed up new system development. OpenAI CEO Sam Altman said that demand for advanced AI infrastructure continues to exceed supply, and that new AI models will require new types of server racks, cooling systems, and power hardware.

A series of reports from Korean and Taiwanese outlets suggests growing pressure across the PC hardware industry as DRAM, NAND, and even NOR Flash prices continue to climb. Korea Economic Daily says NVIDIA, AMD, and other GPU makers are considering dropping some mid-to-high-end gaming cards where memory now makes up an unusually large share of total cost. Taiwanese PC brands, including ASUS, are also said to be looking at ways to reduce memory configurations in upcoming products. TrendForce warns that manufacturers may shift more production toward lower-margin models or raise prices across the board to cope with higher component costs. With most modern PCs, laptops, consoles, tablets, and phones now shipping with at least 16 GB of RAM, any further price spikes or supply shortages could force major brands to cut orders and increase retail pricing. Commercial Times reports that memory alone could add nearly NT$3,000 (about US $96) to basic office PCs next year.

According to Commercial Times, due to rising memory prices some motherboard makers and notebook ODMs vendors have paused new motherboard development or mass production. DRAM, NAND, and NOR Flash supply is tightening at the same time. DDR4 is becoming especially scarce as suppliers accelerate phase-outs and move mature-node capacity toward HBM and DDR5. WJ Capital Perspective estimates a DDR4 shortfall of around 70,000 wafers by the end of 2025, with 2026 unlikely to recover fully. NOR Flash demand is rising as well, driven by AI servers. Commercial Times says NVIDIA's GB200 NVL72 system already uses more than US $600 of NOR per rack, possibly reaching $900 within two years. This comes right after earlier reports that AMD is preparing GPU price increases and Samsung is raising memory prices by up to 60% due to tight supply.

ASUS has cautioned that it might need to raise prices for its PC solutions across the board if component shortages, such as DRAM and NAND flash, continue for a while longer. Speaking for Taiwan's Liberty Times, ASUS co-CEO has made remarks that: "For distribution channel pricing, ASUS will take into account costs, distribution partners and consumer demand, and then adjust the product mix and adjust product prices accordingly." This means that the increase in industry costs will be reflected in the final pricing that ASUS offers its customers. For example, if the price of DRAM increases, a PC with 16, 32, or 64 GB of DDR5 memory will likely see a price increase of a few hundred US dollars.

We recently reported that DRAM prices have surged 171.8% year-over-year, making this "commodity" one of the most valuable assets for applications ranging from data centers to home builds. Even more concerning is the fact that vendors can expect only 70% order fill rate, and smaller OEMs and channel distributors have been told to expect only 35-40% fulfillment through the first quarter of 2026. In addition to DRAM, NAND flash storage has also doubled in the past six months. Phison CEO Khein-Seng Pua recently noted that the price of a 1 Terabit TLC NAND jumped from $4.80 in July 2025 to $10.70 in November 2025, marking an increase of over 100% in less than six months. Other types of NAND flash, such as MLC and QLC, have also seen spot prices double.

TSMC is set to begin construction of its upcoming A14 fab plant in Taichung, Taiwan, as the company has received all necessary permits, Focus Taiwan reports. The project will include the main fab, a central utility plant for power and water management, and several support buildings. The A14 (1.4 nm) node is TSMC's next major process milestone after the 2 nm node and is expected to deliver up to 15% higher performance at the same power. Alternatively, at similar speeds to the 2 nm process, the 1.4 nm node will use 30% less power. The A14 node is designed mainly for AI acceleration and high-performance computing chips, with TSMC targeting mass production in the second half of 2028. The new fab is estimated to be a $49 billion single-site investment from TSMC, one of its largest endeavors.

TSMC recently confirmed that its most advanced N2 (2 nm) node will begin volume production by the end of the year, with a rapid increase planned through 2026 in Taiwan. The announcement follows record Q3 revenue of $33.1 billion, up 41% year-over-year, results mainly driven by strong demand for advanced nodes. Current 3 nm (N3), 5 nm (N5), and 7 nm (N7) nodes remain popular due to increased market demand, with N3 shipments accounting for 23% of wafer revenue, N5 for 37%, and N7 for 14%.

Intel might be the key to re-realizing the American dream of advanced chip production and packaging on U.S. soil. Under the Trump administration, the United States has been making significant efforts to establish leading-edge chip manufacturing domestically. Achieving this goal is challenging for several reasons. In the race for the most advanced silicon, only a few major veteran players remain: Intel, Samsung, and TSMC. Among these, Taiwan-based TSMC has consistently overcome obstacles to become the leader in the semiconductor industry. Through a strategic approach to semiconductor development, TSMC has excelled in both leading-edge node production and advanced chip packaging, enhancing performance.

Historically, Intel has faced difficulties with leading-edge semiconductor node production, even outsourcing some chip manufacturing to TSMC. However, there is a significant opportunity for Intel to not only produce silicon but also establish itself as a major packaging partner for many manufacturers, including TSMC. TSMC's facilities in Arizona address the issue of USA-based manufacturing only partially. While TSMC's Arizona Fab 21 produces 4 nm wafers, these wafers must be sent back to Taiwan for packaging, disrupting the sovereign supply chain that is crucial for domestic manufacturing. Addressing these issues could present a good opportunity for Intel, even if Team Blue doesn't manufacture the underlying silicon.

Just three days ago, NVIDIA and TSMC celebrated the production of the first "Blackwell" GPU at TSMC's Fab 21 complex in Phoenix, Arizona. This marked a significant milestone for both companies, as it signifies the creation of the world's most powerful silicon on U.S. soil. However, the production phase is only part of the story, as the chip benefits not only from an advanced node but also from TSMC's advanced packaging systems. Specifically, the Blackwell family of GPUs utilizes TSMC's Chip on Wafer on Substrate (CoWoS) packaging system.

For CoWoS access, TSMC must send these chips back to the Taiwanese facilities overseas and house the raw die in a CoWoS-S configuration with silicon interposer. This establishes a supply chain dependency on Taiwanese facilities, which remain essential for completing the production of advanced silicon. As a result, the chip isn't entirely manufactured on American soil, indicating that some issues still need to be resolved. TSMC produces NVIDIA's Blackwell chips using a custom 4N 4 nm class node specifically designed for NVIDIA. This die is paired with eight stacks of HBM3E, all housed within a CoWoS-S package.

TSMC's strategy to keep advanced manufacturing primarily in Taiwan might be shifting, as indicated by recent comments from CEO C.C. Wei. During the Q3 earnings call, Wei stated, "We are preparing to upgrade our technologies faster... to more advanced process technologies in Arizona, given the strong AI-related demand from our customers." Previously, TSMC has emphasized that its latest nodes would be developed in Taiwan first, with Arizona following. Currently, TSMC produces N4 4 nm technology at its Arizona facilities and had planned to introduce the N3 node in 2028. However, this plan might change in favor of developing more advanced N2 or even A16 nodes in the coming years, as Taiwanese facilities advance to the A16 and A14 nodes first. The U.S. fabs will remain one node generation behind the Taiwanese fabs, rather than several generations as initially planned.

The push to accelerate advanced node deployment is largely driven by demand from U.S. customers. TSMC recently confirmed that its most advanced N2 node will begin volume production by the end of the year, with a rapid increase planned through 2026 in Taiwan. This marks the second 2 nm-class node to reach high-volume manufacturing, following Intel's 18A node. The competition from Intel's 18A and the upcoming 14A ramp may have led TSMC to reconsider its strategy for bringing advanced manufacturing to its U.S. facilities. While TSMC plans to add five more facilities at its Fab 21 site, it is yet to be confirmed which nodes will be developed there. With an accelerated timeline, it seems only a matter of time before we see the first Angstrom-era nodes produced on American soil.

TSMC has achieved a significant milestone with the first wafer exposure using its N2 process technology at the Kaohsiung Fab 22 facility. This sets the stage for volume manufacturing by late 2025. The accomplishment demonstrates readiness to deliver next-generation semiconductors as Phase 1 nears completion and Phase 2 equipment installation progresses following its groundbreaking in August. Initially planned in 2021 for mature node production at 7 nm and 28 nm, the facility underwent a major transformation to support cutting-edge processes due to rising demand for advanced mobile, AI, and HPC chips. This NT$1.5 trillion investment, spread across six phases, will position southern Taiwan as a hub for sub-2 nm chip manufacturing.

The production roadmap outlines TSMC's strategic approach to meeting the increasing demand from the artificial intelligence and high-performance computing sectors. Phase 2 aims for volume production by the second quarter of 2026, aligning well with the industry's traditional third-quarter procurement cycle when major clients finalize orders for flagship products. Subsequent phases will introduce progressively advanced technologies, including the A16 variant with backside power delivery and eventually the A14 node using gate-all-around nanosheet transistors by 2028. This timeline significantly outpaces international competitors, particularly TSMC's Arizona operations, which are experiencing delays in reaching similar milestones. The Kaohsiung complex is part of TSMC manufacturing network that includes Hsinchu's Fab 20 and the upcoming Taichung Fab 25, providing flexibility to scale production as global semiconductor demands grow.

The U.S. government is considering imposing massive tariffs on semiconductor products unless manufacturers can match those volumes with chips produced in the United States. Under the current draft, companies that do not match foreign-sourced chip volumes with domestic output over time could face tariffs that, in some versions, reach 100%. The idea is an act of forceful action to speed up onshore capacity and reduce strategic dependence on overseas foundries, but they also admit that the plan faces serious practical hurdles. The core concept is simple to draft on paper, but hard to implement. For every unit brought into the U.S. from an overseas plant, such as TSMC in Taiwan and Samsung in South Korea, a corresponding unit would need to be produced domestically within a specified timeframe, or tariffs would be applied. This means that for every batch of chips NVIDIA, AMD, Intel, or Apple produce overseas, the same chip would need to be produced in the U.S.

Counting and comparing chip "units" is extremely complicated. Every semiconductor design differs wildly by complexity, cost, and function, and the typical semiconductor supply chain moves wafers, packages, and finished devices across several countries before a product reaches consumers. Even TSMC, the very maker of the chip, sometimes has a hard time confirming if the designs it makes are on the U.S. sanctions list. Like it experienced with Huawei, TSMC couldn't determine that a third-party shell company was ordering Ascend accelerator IP on behalf of Huawei—which sits on the U.S. blacklist—even with advanced detection tools. Officials are reportedly considering transitional credits and carve-outs for companies that commit to building U.S. fabs, which would allow them to import while new plants come online. Even with those credits, a question arises about how to compare a million small mobile processors against a million high-performance data center accelerators.

TSMC is the only semiconductor maker to move its nodes at incredible speeds from design to mass production. According to Taiwanese media outlet Taipei Times, TSMC is expected to lay the groundwork for Fab 25, a multiphase manufacturing complex planned in Taichung's Central Taiwan Science Park. Company officials say construction will begin in the Q4 of this year and that the first of four planned factories could be finished by 2027, setting the stage for commercial A14 chip production by the end of 2028. The initial funding is large but only as TSMC has prepared 500 billion NTD, roughly $16.38 billion, for the first stage, with many more billions expected as the full site is built out. This push is intended to bring capacity online sooner than anticipated, so customers don't experience any delays.

TSMC executives have denied recent rumors that packaging or U.S. projects are forcing delays to the Taiwan buildout. Technical and operational planning is designed to reduce risk and accelerate volume. TSMC plans to seed A14 production earlier at Phases 3 and 4 of Fab 20 near its Hsinchu R&D campus, while Fab 25 will eventually serve as a dedicated high-volume home for the node. The initial A14 rollout will follow conventional process improvements, with a succeeding A14 Plus or A14P revision expected to introduce additional improvements. Still, advanced packaging remains the tightest constraint for the industry, and a lasting balance between supply and demand is unlikely to appear anytime soon.