Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/tsmc-widens-gap-with-samsung-by-expanding-2nm-and-3nm-output-by-20.25032/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2031070
            [XFI] => 1060170
        )

    [wordpress] => /var/www/html
)

TSMC Widens Gap with Samsung by Expanding 2nm and 3nm Output by 20%

Daniel Nenni

Founder
Staff member

The move widens TSMC’s lead over Samsung, as Intel also presses into the foundry market

TSMC increases 2nm and 3nm production by 20%, widening Gap with Samsung in advanced chips.


TSMC increases 2nm and 3nm production by 20%, widening Gap with Samsung in advanced chips.

Taiwan’s TSMC, the world’s top foundry company, will expand its advanced process production capacity by 20% this year, further cementing its grip on the number one foundry position.

Taiwan’s TSMC, the undisputed dominant force in the global semiconductor contract manufacturing (foundry) market, is now set to expand its cutting-edge 2-nanometer (nm) and 3nm production capacity by 20%. This signals that TSMC’s 2nm process yield has already reached a stage of stabilization.

As TSMC scales up output in advanced processes, Samsung Electronics’ foundry standing is set to shrink further. The goal set by Samsung Electronics Chairman Lee Jae-yong to achieve global leadership in the non-memory sector including foundry by 2030 is also growing more distant.

According to market research firm Counterpoint Research, TSMC held a market share of more than 70% last year, while Samsung Electronics stood at 6% to 7%.

For Samsung’s foundry division, the 4nm node is currently a mature process, while the company is still focused on improving yields for its 2nm node. On top of that, Samsung is carrying a range of internal risks, including delayed mass production in the sub-1nm range, the burden of securing customers, and labor union issues.

The gap is expected to widen further. TSMC plans to begin mass production at the 1nm class from 2027, while Samsung Electronics has pushed back its own target for sub-1nm mass production by approximately two years from an original goal of 2027 to 2029.

According to Taiwanese and international media on April 29, TSMC has decided to expand wafer production capacity for its 2nm and 3nm processes by approximately 20% by the end of 2026, in response to growing demand for artificial intelligence (AI) semiconductors.

TSMC plans to expand its monthly 3nm process production capacity to approximately 180,000 wafers and aims to scale its 2nm capacity up to around 100,000 wafers per month. Industry observers believe TSMC has already secured a stable yield of 80% to 90% across both its 2nm and 3nm processes.

The expansion is attributed to growing AI semiconductor orders from major big tech customers including Nvidia, AMD, and Apple. Analysts note that as demand for data center graphics processing units (GPUs) and high-performance AI chips rapidly expands, structural demand centered on advanced fine-pitch processes is intensifying.

As a result, major big tech companies are reportedly required to wait at least six months or more to place foundry orders with TSMC. This is also cited as the reason why some companies, including Tesla and Apple, have recently turned to Samsung’s foundry for advanced chip manufacturing.

However, if TSMC increases its production volume, big tech companies that had been directing orders toward Samsung are likely to reverse course and return to TSMC. Given the nature of the foundry business, where chip designs must be shared and stable production is essential, switching manufacturers is not common practice. The market expects this capacity expansion to serve as a turning point that further consolidates TSMC’s dominance in advanced processes.

Meanwhile, Intel’s re-entry into the foundry market is also acting as a negative variable for Samsung’s foundry division. Intel, backed by U.S. government support, is currently in the process of bringing in equipment for mass production of its 14A (1.4nm class) process next year.

Intel is also strengthening its competitiveness by recruiting former Samsung Electronics personnel and participating in Tesla’s AI chip project “Terafab.”

Ultimately, as the industry enters the AI semiconductor era, the global foundry market is shaping into a complex three-way dynamic: TSMC pushing an ever-widening lead, Intel closing in from behind, and Samsung scrambling to hold its ground.

Samsung Electronics, for its part, is seeking to secure customers by leading with its proven 4nm FinFET process, which offers verified yield and stability, rather than competing head-on in the cutting-edge 2nm space. This is interpreted as a strategy to broaden the foundry order base by targeting not only customers demanding the most advanced processes but also those who place greater weight on mass production stability and cost efficiency.

The company explained that the competitiveness of its 4nm process lies in its high maturity and stable yield, backed by more than six years of mass production experience. It also emphasized that the platform can serve a wide range of applications, including AI accelerators, sixth-generation high-bandwidth memory (HBM4) base dies, automotive semiconductors, and radio frequency (RF) chips.

However, Samsung Electronics is currently facing an unexpected setback in the form of a threatened general strike by its labor union. Given the nature of the foundry business, once customer trust is lost, it takes considerable time to recover.

Kim Dae-jong, a professor at the Sejong University College of Business Administration, analyzed that “without simultaneous progress in yield stabilization, customer ecosystem expansion, and organizational structural reform, the recovery of competitiveness will not be easy.”

 
Intel is also strengthening its competitiveness by recruiting former Samsung Electronics personnel and participating in Tesla’s AI chip project “Terafab.”

Intel is helping Elon Musk build a foundry operation that will compete directly with Intel’s own product division and foundry services division. How can that possibly “strengthen” Intel?
 

The move widens TSMC’s lead over Samsung, as Intel also presses into the foundry market

TSMC increases 2nm and 3nm production by 20%, widening Gap with Samsung in advanced chips.


TSMC increases 2nm and 3nm production by 20%, widening Gap with Samsung in advanced chips.

Taiwan’s TSMC, the world’s top foundry company, will expand its advanced process production capacity by 20% this year, further cementing its grip on the number one foundry position.

Taiwan’s TSMC, the undisputed dominant force in the global semiconductor contract manufacturing (foundry) market, is now set to expand its cutting-edge 2-nanometer (nm) and 3nm production capacity by 20%. This signals that TSMC’s 2nm process yield has already reached a stage of stabilization.

As TSMC scales up output in advanced processes, Samsung Electronics’ foundry standing is set to shrink further. The goal set by Samsung Electronics Chairman Lee Jae-yong to achieve global leadership in the non-memory sector including foundry by 2030 is also growing more distant.

According to market research firm Counterpoint Research, TSMC held a market share of more than 70% last year, while Samsung Electronics stood at 6% to 7%.

For Samsung’s foundry division, the 4nm node is currently a mature process, while the company is still focused on improving yields for its 2nm node. On top of that, Samsung is carrying a range of internal risks, including delayed mass production in the sub-1nm range, the burden of securing customers, and labor union issues.

The gap is expected to widen further. TSMC plans to begin mass production at the 1nm class from 2027, while Samsung Electronics has pushed back its own target for sub-1nm mass production by approximately two years from an original goal of 2027 to 2029.

According to Taiwanese and international media on April 29, TSMC has decided to expand wafer production capacity for its 2nm and 3nm processes by approximately 20% by the end of 2026, in response to growing demand for artificial intelligence (AI) semiconductors.

TSMC plans to expand its monthly 3nm process production capacity to approximately 180,000 wafers and aims to scale its 2nm capacity up to around 100,000 wafers per month. Industry observers believe TSMC has already secured a stable yield of 80% to 90% across both its 2nm and 3nm processes.

The expansion is attributed to growing AI semiconductor orders from major big tech customers including Nvidia, AMD, and Apple. Analysts note that as demand for data center graphics processing units (GPUs) and high-performance AI chips rapidly expands, structural demand centered on advanced fine-pitch processes is intensifying.

As a result, major big tech companies are reportedly required to wait at least six months or more to place foundry orders with TSMC. This is also cited as the reason why some companies, including Tesla and Apple, have recently turned to Samsung’s foundry for advanced chip manufacturing.

However, if TSMC increases its production volume, big tech companies that had been directing orders toward Samsung are likely to reverse course and return to TSMC. Given the nature of the foundry business, where chip designs must be shared and stable production is essential, switching manufacturers is not common practice. The market expects this capacity expansion to serve as a turning point that further consolidates TSMC’s dominance in advanced processes.

Meanwhile, Intel’s re-entry into the foundry market is also acting as a negative variable for Samsung’s foundry division. Intel, backed by U.S. government support, is currently in the process of bringing in equipment for mass production of its 14A (1.4nm class) process next year.

Intel is also strengthening its competitiveness by recruiting former Samsung Electronics personnel and participating in Tesla’s AI chip project “Terafab.”

Ultimately, as the industry enters the AI semiconductor era, the global foundry market is shaping into a complex three-way dynamic: TSMC pushing an ever-widening lead, Intel closing in from behind, and Samsung scrambling to hold its ground.

Samsung Electronics, for its part, is seeking to secure customers by leading with its proven 4nm FinFET process, which offers verified yield and stability, rather than competing head-on in the cutting-edge 2nm space. This is interpreted as a strategy to broaden the foundry order base by targeting not only customers demanding the most advanced processes but also those who place greater weight on mass production stability and cost efficiency.

The company explained that the competitiveness of its 4nm process lies in its high maturity and stable yield, backed by more than six years of mass production experience. It also emphasized that the platform can serve a wide range of applications, including AI accelerators, sixth-generation high-bandwidth memory (HBM4) base dies, automotive semiconductors, and radio frequency (RF) chips.

However, Samsung Electronics is currently facing an unexpected setback in the form of a threatened general strike by its labor union. Given the nature of the foundry business, once customer trust is lost, it takes considerable time to recover.

Kim Dae-jong, a professor at the Sejong University College of Business Administration, analyzed that “without simultaneous progress in yield stabilization, customer ecosystem expansion, and organizational structural reform, the recovery of competitiveness will not be easy.”


They do seem very busy of late , I assume they running high capacity usage on what I see in our system.

Usually a good sign for their revenue figures 🤣🤣🤣
 
Back
Top