News TSMC Places Massive EUV Tools Order to Boost Capacity

[Admin]

TSMC reportedly buys over a dozen of ASML Twinscan NXE EUV machines

TSMC Places Massive EUV Tools Order to Boost Capacity : Read more

 

[neojack]

i always wonder why ASML does not inflate the price of their tools over the top.
they are unique, they could sell them 1B each or something. sky is the limit. they have a monoply on a technology everyone wants.

or maybe their management is more interested into the greater good and not just capitalism ?
not complaining for sure. don't want having to pay 5000$ for a CPU.

 

[InvalidError]

i always wonder why ASML does not inflate the price of their tools over the top.
they are unique, they could sell them 1B each or something. sky is the limit. they have a monoply on a technology everyone wants.

ASML cannot sell equipment to its customers at stupid markups because most of its clients are also R&D partners, investors and in many cases, they are also on the board of directors. In other words, foundries are basically outsourcing their equipment R&D to ASML instead of individually reinventing the photolithography wheel in exchange for fair pricing. Intel divested itself of some of its ASML holdings in 2018 and still has a ~3% stake in it.

Intel also divested part of its ASML stake back in 2012, which coincides with the beginning of Intel's process slip-ups.

 

[neojack]

i didn't know that, very interesting ! take a like

maybe we can deduce that intel's moving away from TSMC also costed them R&D efforts. they turned their backs from ASML because of their hubris

2014 intel : "we are the best fab in the world and we do not need subcontractors like ASML". "our 14nm is the best !"
2020 intel : still on 14nm, struggling with their 10nm.

 

[GoalTorrent]

TSMC reportedly buys over a dozen of ASML Twinscan NXE EUV machines

TSMC Places Massive EUV Tools Order to Boost Capacity : Read more

What is the status of the photoresist being used at TSMC or any other company using EUVL? There are several in the running, but almost nothing published on who seems to have gained the place in the fab.

 

[jkflipflop98]

i didn't know that, very interesting ! take a like

maybe we can deduce that intel's moving away from TSMC also costed them R&D efforts. they turned their backs from ASML because of their hubris

2014 intel : "we are the best fab in the world and we do not need subcontractors like ASML". "our 14nm is the best !"
2020 intel : still on 14nm, struggling with their 10nm.

Uh, what? That makes no sense. Intel uses plenty of everyone's tools. Including ASML.

 

[InvalidError]

Uh, what? That makes no sense. Intel uses plenty of everyone's tools. Including ASML.

While Intel may use ASML stuff, the reason Intel is in its current predicament is because it thought it could leapfrog ASML by aggressively refining the equipment it already had and it has turned into a mostly unmitigated failure so far. Had Intel retained the same level of partnership it had with ASML before 2012, maybe it wouldn't have fallen as far behind.

 

[jkflipflop98]

While Intel may use ASML stuff, the reason Intel is in its current predicament is because it thought it could leapfrog ASML by aggressively refining the equipment it already had and it has turned into a mostly unmitigated failure so far. Had Intel retained the same level of partnership it had with ASML before 2012, maybe it wouldn't have fallen as far behind.

Again, that makes no sense. Guess who had these EUV tools before TSMC? Intel. Lithography is always being researched and worked on. It's the hard part of making chips. It's not like all that research is going to waste. You can apply the same tricks you learned at 190nm to the newer wave.

 

[InvalidError]

Again, that makes no sense. Guess who had these EUV tools before TSMC? Intel.

AFAIK, Intel only plans to start using EUV for 7nm in 2021, everything it currently has in production is still DUV with increasingly fancy multi-patterned computational lithography.

 

[jkflipflop98]

You heard wrong, dude.

 

[InvalidError]

You heard wrong, dude.

If you do a simple search for Intel and EUV, the only results you get are related to 7nm. Even Tiger Lake on 10nm+ is still quad-patterned DUV. The only EUV "production" Intel currently has is R&D to (hopefully) figure 7nm out before the remainder of its EUV equipment rolls in and the six months slip on 7nm launches suggests it may not be going well.