Skip to content
World clockEU--:--UK--:--USA--:--CN--:--PLDEFRIT中文EN

portal about AI and technologyevents · analysis · interviews · technical background

Search
LIVE
›

Samsung orders equipment for 6,000 more 2nm wafers a month at Taylor

Samsung has ordered equipment to add 6,000 2nm wafers a month at its Taylor, Texas fab, ZDNet Korea reported, aiming for a 20,000-wafer monthly maximum by the second half of 2027 for customers including Tesla.

TechnologyAnalysisRachel NwosuPublished: 6 November 20264 min readSources 24
Samsung orders equipment for 6,000 more 2nm wafers a month at Taylor

The 6,000-wafer order

Samsung has placed an equipment order that will let its Taylor 1 plant produce 6,000 wafers a month of 2nm chips, ZDNet Korea reported. Taylor 1 is expected to become operational by the end of this year and will use extreme ultraviolet lithography machines, according to the report. The order builds on earlier procurement aimed at 10,000 wafers a month. Together, the earlier procurement and the new order cover 16,000 wafers a month. ZDNet Korea's sources said pilot production capacity would close the remaining 4,000-wafer gap to the site's 20,000-wafer monthly maximum by the second half of 2027.

Capacity and Tesla

The Taylor fab is equipped for 50,000 wafers a month, Electronics Weekly reported, so the 20,000-wafer maximum for the second half of 2027 would be 40% of that capacity. The 20,000 figure is a ceiling, not current output. The fab's 2nm capacity is fully booked with Tesla, Broadcom and Arm, according to Electronics Weekly. Digitimes reported in September that the booking was complete before volume production began.

Tesla remains the key customer at Taylor, Electronics Weekly reported, having signed a $16.5 billion supply deal with Samsung. Samsung has started trial production of Tesla's AI5 chip at Taylor, Seoul Economic Daily reported, with prototype chips running on several lines as part of calibration. Digitimes reported that the work moved the fab into 2nm wafer production and yield verification. Samsung intends to begin supplying production AI5 chips in 2027, with volume in mid-2027, Electronics Weekly reported.

Yield recovery

Yield estimates for Samsung's 2nm process have varied by source and date. TechTimes reported that analysts put yields at roughly 50% when Exynos 2600 mass production began in September 2025. Sources cited by Korea Economic Daily said yields crossed 60% at a peak in the first quarter of 2026, climbing from roughly 20% in late 2025. Busan Ilbo reported in April 2026 that yields were in the mid-50s. TSMC was reportedly at 60% to 70% at its own 2nm node. The reports point to improvement from about 20% in late 2025 to above 60% at the early 2026 peak, though the April estimate shows the level was not steady.

Timeline and next fab

Samsung announced in 2021 that it would invest more than $17 billion in Taylor to build a new foundry plant, Chosun Biz reported. In its second-quarter earnings release, the company said, "Taylor Fab 1 is proceeding with preparations for operation in 2026 as planned." Samsung plans to start construction of Taylor 2 at the end of this year and begin mass production in 2030, Chosun Biz reported. The second fab is targeted to use a 1.4nm process, Electronics Weekly reported. The company says expanding advanced-process capacity is necessary because of growing AI demand.

Comments 0

Sources

24
  1. 01Samsung Orders Equipment For 6,000 More 2nm Wafers A Month At Its Taylor, Texas Fab, Putting 20,000 Within Reach By The Second Half Of 2027EN
  2. 02Tesla's AI5 Chip Enters Trial Production As Samsung Gets the Taylor Fab RunningEN
  3. 03Samsung reportedly starts Tesla chip trial production at Taylor fabEN
  4. 04Samsung's Taylor fab running prototypes of Tesla's 2nm A15 chipEN
  5. 05Samsung's Taylor fab reportedly fully booked before 2nm production beginsEN
  6. 06Texas showdown heats up: Samsung counters TSMC's advance with $22.84m investmentEN
  7. 07Tesla AI5 Locks In Samsung 2nm at Taylor, Flipping the Node AssumptionEN
  8. 08Low pay dampens Korea Samsung dispatch to US Taylor fab before ramp-upEN
  9. 09Memory boom carries Samsung to another bumper quarterEN
  10. 10Samsung's Texas Timeline Stretches to 2030 While Wall Street Fixates on October 29EN
  11. 11U.S. startup secures $9 million to scale production of 22.33%-efficient perovskite solar modulesEN
  12. 12 Grab a huge 44% saving on a 4K 43-inch Samsung smart monitor, now under $280 EN
  13. 13Samsung’s Galaxy Z Fold 8 is $300 off ahead of Black FridayEN
  14. 14Aldi mischt Postpaid-Markt auf: Neue Tarife mit Samsung und AppleEN
  15. 15Claimlane bags €5.6M to turn product claims into business valueEN
  16. 16The Download: Texas’s fragile AI boom and energy policy shiftsEN
  17. 17Texas was the new epicenter of the AI data center buildout. That might be kaput.EN
  18. 18What does a blue Texas sweep mean for Elon Musk, Tesla & SpaceX?EN
  19. 19Tesla wins trade zone status for Texas Megapacks, but China tariffs stayEN
  20. 20Samsung suffers its worst 5G quarter but knows it can outlast rivalsEN
  21. 21Merge builds product and business tools with OpenAI technologyEN
  22. 22Flipsnack grows product signups with conversion optimizationEN
  23. 23Li Auto's Mega faces Samsung OLED screen shortage as deliveries reboundEN
  24. 24Astranis Upgrades Falcon 9 Launch Contract to Dedicated Mission for Entire Block 4 Production Run of SatellitesEN

All figures and quotations in this text come from the sources listed below.

Content prepared by the editorial team with AI assistance.

Rachel Nwosu

Rachel Nwosu

AI, models and technology

Rachel Nwosu covers AI, models and technology for FLASH24, working from public model documentation, benchmark releases and repository histories rather than press summaries, and she skips announcements that arrive without reproducible numbers. She checks training-data claims against dataset cards and reruns reported metrics where code is available. She spends much of her week interviewing researchers and engineers, tracking model launch calendars, and comparing vendor benchmarks with independent evaluations. Outside the desk she runs 3D printers, restores old computers, and tests how models learn from internet junk. She does not publish benchmark figures she cannot trace to a source.

Newsroom →

Comments

0
  1. No comments yet — be the first.

Write a comment

Comments are public. We do not publish abuse, spam or advertising.