Semiwiki on LinkedIn Semiwiki on Facebook Semiwiki on Twitter Register / Log in 
Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/q-a-with-tsmc-on-next-gen-foundry.20696/
        )

    [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] => 2021770
            [XFI] => 1050270
        )

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

Q&A With TSMC on Next-Gen Foundry

Daniel Nenni

Daniel Nenni

Admin
Staff member
Discussions with Dr. Kevin Zhang, SVP
Dr. Ian Cutress

Over the past year, I’ve had the chance to quiz every major foundry about their plans for the future. The latest in this series is TSMC, as part of their recent announcements around their roadmaps and upcoming process node technologies. As the world’s leading manufacturer of leading edge and EUV-based logic, as well as packaging the vast majority of the big AI chips, the pressure on TSMC to execute at speed and scale has only increased over the last few years. As part of their commitments to next-generation technologies, the company announced its A16 process node, Super Power Rail (the marketing name for its backside power delivery), ventures into co-packaged optics, and Silicon-on-Wafer technology, leveraging a demand for bigger substrates for the biggest chips. Here’s my latest video on these announcements.

 
I had heard rumors and hoped that TSMC would build packaging in AZ. They certainly have the room for it.

Q: There are also discussions about you expanding worldwide production with multiple new fabs around the world announced. How is that progressing?

A: It's progressing very well. And they're very fast too. If you look at our manufacturing footprint, it has expanded quite significantly. Just the over last few years we are expanding quickly in our Arizona site. We built the first fab, focusing on 4nm, which we're going to move into production next year. We are also building the second fab there, and we announced a third fab there. We're going to continue to bring the most advanced node to North America, as that's where our largest customer base is. Right from 4nm to 3nm, down to 2nm, or even A16 in the future. It’s very exciting!

At the same time, we're expanding our specialty technology in both Japan and Europe. In Japan, the Kumamoto project has gone well, and we're going to produce it in the second half of this year. We're going to bring most advanced MCU embedded non-volatile memory, which is very important for the auto industry here in Europe,

Q: Does that extend to packaging in any way?
A: We are evaluating the option, but at the same time, I think right now we are working closely with our partners to bring up the manufacturing capabilities here in the US.
 
Daniel Nenni said:
We're going to continue to bring the most advanced node to North America, as that's where our largest customer base is. Right from 4nm to 3nm, down to 2nm, or even A16 in the future. It’s very exciting!
Click to expand...
Perhaps I haven't been paying close enough attention lately, but wasn't TSMC's strategy to reserve the most advanced node production for Taiwan? Is this essentially an announcement, or did I miss the announcement that Arizona would get the latest node too?
 
blueone said:
Perhaps I haven't been paying close enough attention lately, but wasn't TSMC's strategy to reserve the most advanced node production for Taiwan? Is this essentially an announcement, or did I miss the announcement that Arizona would get the latest node too?
Click to expand...

Kevin says in the interview that A16 will be in Taiwan. N3 is still in Taiwan but will be coming to AZ. I have not heard of plans for N2 in AZ. TSMC had originally planned for N5/4 in AZ but N3 demand is very high and the fabs and equipment are the same (90%) so the N3 in AZ decision was easy. They will need to build new fabs for N2 and A16. That is my understanding.

Q: We're expecting this all to be manufactured in Taiwan?
A: A16 will start in Taiwan.
 
Kevin also talks about BSPD:

A16 is a major technology enhancement, and is revolutionary in terms of bringing further power and performance to future high-performance applications, especially those targeting HPC and AI. A16 features nanosheet transistors, which are the industry-leading and the most advanced transistor architecture.

At the same time we have added a very innovative backside power rail design. This backside power rail design allows the customer to move the power supply routing from the front to the back, opening up the space to enhance the performance and at the same time improve the power supply. Our approach is very different from the conventional design for BSPDN - in the conventional backside power rail, you just a drill hole to connect the backside metal to the front side of metal. In doing so you burn space, and you have to enlarge the footprint of the library cell. But in our design, we have very innovative approach - we move the contact or transistor, the source of the transistor to the back without changing the footprint of the library cells. So this clever way allows us to maintain the footprint and provide the maximum flexibility to our customers.

So we are targeting the second half of 2026 to go into production for A16 for lead customers.
 
1722037746734.png

 
Last edited by a moderator:
All the phases for the TSMC fab in Az are small scale 20K/month compared to the GigaFab of multiphase with >> 100k/month in Taiwan. They are neither cost nor manufacturing efficient for high mix low volume production. The N4 coming online in 2025 and N3/N2 2-4 years later are a good 5 years and a trickle compared to the volume in Taiwan. The need and product mix need is a mess as most demand has moved to the N+1 node and N+2.

Also in small fabs there are always a few steps where one tool is almost capable of supplying all the needs of a phase that make the gigafab for more efficient as well as productive when some thing happens. A small fab like 21 makes expensive choices of 2 tools or maybe 3 for redundancy and run them at very low utilization or just one and when that tool or one of two go down lose 1/2 or 100% of capacity. Makes for operational nightmare for new product and high mix low volume.

The giga fab is the way to go and the small fab in US is all politics not business sense or continuity. If TSMC is serious they need far larger scale.

The real foundry is in Taiwan!
 
AZ update:

About TSMC Arizona

In the city of Phoenix, TSMC Arizona will represent the world’s most advanced semiconductor technology in the United States. In a historic announcement, in May 2020, TSMC shared its plans to invest $12B in Phoenix, Arizona – building an advanced semiconductor manufacturing fabrication. In...
semiwiki.com
 
Daniel Nenni said:
AZ update:

About TSMC Arizona

In the city of Phoenix, TSMC Arizona will represent the world’s most advanced semiconductor technology in the United States. In a historic announcement, in May 2020, TSMC shared its plans to invest $12B in Phoenix, Arizona – building an advanced semiconductor manufacturing fabrication. In...
semiwiki.com
Click to expand...
To go big and international would be to put the majority of a node in a place outside of Taiwan, till than all politics. Even with N4, N3 and N2 in Az. If Taiwan was to go down that fab overseas will be of little value
 
Daniel Nenni said:
It is careful balancing act between world politics and the Silicon Shield. They are doing quite well in my opinion.
Click to expand...
I believe it was planned to be careful balancing between geopolitical conflicts and silicon shield at beginning, but the foundry culture conflict between Taiwan and US is a big hurdle still to be overcome. More fab automation, remote engineering operation or even smart manufacturing using ML/AI in operations could be solutions, but physical on-site is still necessary for some functions. It is tough and let's see how tsmc manage it. To scale up, Arizona fab needs to prove the manufacturing efficiency and yield are comparable to Taiwan Fab in small volume (20 kwpm) first. For non-government control companies, no one will scale up without considering profitability and long-term sustainability even with near-term incentive. For example, tsmc did not scale up their Washinton Fab in the past 25 years.
 
hskuo said:
Poor Example? It is the only "tsmc" Fab in US before Arizona fab.
Click to expand...
TSMC built the facility in Camas, WA in 1996, which at the time had a population of less than 20,000. It is 14 miles down WA14 from Vancouver, WA (which had and still has a population of less than 200,000), a highway that runs along the Columbia River Gorge, and often suffers from icy roads in the winter. It is at least 90 minutes from Hillsboro, Oregon during commuting hours, where Intel has its fabs, and most of the experienced engineers live. Camas itself consists mostly of large and expensive homes, so living nearby would not be viable for most fab workers; perhaps in the few apartments there are.

Looking at TSMC Washington's website, they're still describing their fab technology in terms of 8 inch wafers and the fab processes in microns. WaferTech, as it used be known, is not at all comparable to what TSMC has down in Arizona, no less Taiwan.

So, yes, I think the Camas facility is a poor example. I also agree with @BruceA, even the Arizona fabs are not comparable to the Taiwanese fabs economically.
 
Last edited:
blueone said:
TSMC built the facility in Camas, WA in 1996, which at the time had a population of less than 20,000. It is 14 miles down WA14 from Vancouver, WA (which had and still has a population of less than 200,000), a highway that runs along the Columbia River Gorge, and often suffers from icy roads in the winter. It at least 90 minutes from Hillsboro, Oregon during commuting hours, where Intel has its fabs, and most of the experienced engineers live. Camas itself consists mostly of large and expensive homes, so living nearby would not be viable for most fab workers; perhaps in the few apartments there are.

Looking at TSMC Washington's website, they're still describing their fab technology in terms of 8 inch wafers and the fab processes in microns. WaferTech, as it used be known, is not at all comparable to what TSMC has down in Arizona, no less Taiwan.

So, yes, I think the Camas facility is a poor example. I also agree with @BruceA, even the Arizona fabs are not comparable to the Taiwanese fabs economically.
Click to expand...
Did you read my sentence clearly? "For example, tsmc did not scale up their Washinton Fab in the past 25 years." If you compare Wafertech with tsmc leading fabs now, it is definitely orange to apple comparison. But if you compared with tsmc Taiwan Fabs 25 years ago, Wafertech positioned like Arizona fab to Taiwan fabs now. Why it did not scale up in the past 25 years?
 
hskuo said:
Did you read my sentence clearly? "For example, tsmc did not scale up their Washinton Fab in the past 25 years." If you compare Wafertech with tsmc leading fabs now, it is definitely orange to apple comparison. But if you compared with tsmc Taiwan Fabs 25 years ago, Wafertech positioned like Arizona fab to Taiwan fabs now. Why it did not scale up in the past 25 years?
Click to expand...
I read your post clearly, and I understood your implication. I just don't buy it. I've been to Camas, and it sure looks like an unlikely place to put a fab you would ever intend to scale. Have you ever been to Camas?
 
hskuo said:
Did you read my sentence clearly? "For example, tsmc did not scale up their Washinton Fab in the past 25 years." If you compare Wafertech with tsmc leading fabs now, it is definitely orange to apple comparison. But if you compared with tsmc Taiwan Fabs 25 years ago, Wafertech positioned like Arizona fab to Taiwan fabs now. Why it did not scale up in the past 25 years?
Click to expand...
TSMC was well regarded 25 years ago but not the absolute powerhouse that it is today. Therefore, it’s more likely to attract talent today than back then. In addition, the Arizona state government and local ecosystem is far more supportive.
 
blueone said:
I read your post clearly, and I understood your implication. I just don't buy it. I've been to Camas, and it sure looks like an unlikely place to put a fab you would ever intend to scale. Have you ever been to Camas?
Click to expand...
Camas is rural but Linear Technology (now ADI) had a fab there. At the time, Gresham, OR also had several Japanese built fabs and Gresham is right across the river from Camas. It may not have been as absurd a choice back then as it seems today.
 
blueone said:
I read your post clearly, and I understood your implication. I just don't buy it. I've been to Camas, and it sure looks like an unlikely place to put a fab you would ever intend to scale. Have you ever been to Camas?
Click to expand...
Thank you for your explanation, it is your opinion and it is not relevant that I had been in Wafertech or not.
 
osnium said:
Camas is rural but Linear Technology (now ADI) had a fab there. At the time, Gresham, OR also had several Japanese built fabs and Gresham is right across the river from Camas. It may not have been as absurd a choice back then as it seems today.
Click to expand...
Gresham is over 20 miles from Camas. Gresham I understand. Camas? Not very much. Not for an operation anyone ever intended to scale.
 
You must log in or register to reply here.
Back
Top