The Bleeding Edge of Semiconductors: A Tale of Three Companies || Peter Zeihan
- Thread starter soAsian
- Start date
hist78
Well-known member
He packed some correct and incorrect or incomplete data to create a 5 minutes video in order to explain a complicated and large topic.
MKWVentures
Moderator
IMO: The reason for the issue is incorrect. Intel was planning EUV, they just kept failing on 10nm and 7nm etc... . Intel got EUV tools before most other companies. And Intel was not way ahead in 2017. Intel didnt get cocky.... they failed to execute (there is a long list of reasons why). IF intel was cocky, its the fact that they thought they could do lots of difficult process changes all at the same time and never gave up when they should have.
Intel is overhyping High NA for PR purposes to show they are committed to getting into the lead. we will know whether High NA is an advantage and when sometime in 2026. When High NA is ready, TSMC will ramp it as will Intel .... IF they both execute. who will be the lead company to have all the painful learning? Ask me at end of 2025 (Hint both companies have well defined backup plans for High NA).
If Intel is making 18A and TSMC is Making N2 in production in volume in 2025, this will be a fun race again. But remember, Intel is dependent on TSMC for chips.... and TSMC makes way more leading edge wafers than Intel. This is true for the next 5+ years
Intel is overhyping High NA for PR purposes to show they are committed to getting into the lead. we will know whether High NA is an advantage and when sometime in 2026. When High NA is ready, TSMC will ramp it as will Intel .... IF they both execute. who will be the lead company to have all the painful learning? Ask me at end of 2025 (Hint both companies have well defined backup plans for High NA).
If Intel is making 18A and TSMC is Making N2 in production in volume in 2025, this will be a fun race again. But remember, Intel is dependent on TSMC for chips.... and TSMC makes way more leading edge wafers than Intel. This is true for the next 5+ years
it feels like he is using the topic to expand into geopolitics.
Not much. Clearly he does not work in the semiconductor industry. Nice try though.
What bunch of buzzwords used incorrectly. Good lord. Stay in your lane! "In 1995, he obtained a BS in political science at Northeast Missouri State University and in 1997 a postgraduate diploma in Asian studies at the University of Otago."
tooLongInEDA
Moderator
I've watched quite a lot of Peter Zeihan's geopolitics stuff. He specialises in big picture, "helicopter view" stuff and can be quite convincing (he certainly speaks with enormous self-confidence). He seems to have a pretty good understanding of things like global oil and energy markets and demographics. Also very positive about the future for the US.
I found this piece over-simplistic and factually incorrect in places - but then YouTube is largely about telling stories, so over-simplification works best. The idea that "all of a sudden" Intel Columbus (Ohio) is going to come on stream with high NA EUV 18A (or smaller) production feels a bit of a stretch right now (having just been delayed). The bit about the difference between DUV and EUV just seemed wrong (ignored multiple patterning and claimed there was some magic "digital stuff" in EUV).
That's made me a little more sceptical about some of his other stuff. But I think worth listening to provided you don't take it all without challenge, however plausible he makes it sound.
I found this piece over-simplistic and factually incorrect in places - but then YouTube is largely about telling stories, so over-simplification works best. The idea that "all of a sudden" Intel Columbus (Ohio) is going to come on stream with high NA EUV 18A (or smaller) production feels a bit of a stretch right now (having just been delayed). The bit about the difference between DUV and EUV just seemed wrong (ignored multiple patterning and claimed there was some magic "digital stuff" in EUV).
That's made me a little more sceptical about some of his other stuff. But I think worth listening to provided you don't take it all without challenge, however plausible he makes it sound.
What's missing from the story: 157i
Intel invested massively into new tools, but then the cycle caught up to them, and they had no courage to tank the losses.
In the meanwhile, EUV has matured during the time span for which 157i was originally expected to hold.
I think Intel did not expect phase correcting masks, and very advanced multipatterning to prolong 193i life for so long.
Intel invested massively into new tools, but then the cycle caught up to them, and they had no courage to tank the losses.
In the meanwhile, EUV has matured during the time span for which 157i was originally expected to hold.
I think Intel did not expect phase correcting masks, and very advanced multipatterning to prolong 193i life for so long.
MKWVentures
Moderator
Ah 157nm ..... fun times! I remember when that all fell apart and the term "intrinsic birefringence" became my favorite hot topic. That was way back when Intel was the leader in process nodes.
Did 157nm immersion happen before. I thought 157nm died before arrival and 193i filled the gap before EUV ready.
157nm, good all times. Intel declaring at SPIE the needs of tons of CaF2 crystals for the hundreds of steppers they were going to buy just 3 weeks before ditching everything. Schott ever so glad of keeping with the plan of waiting until the first full demonstration of the stepper before boosting capacity The skwed pupil of the first tool..... A lot of fond memories.
157i that a nice one !!!!!!
I would really like to see the CaF2 optics getting near water
Even better was the story of the higher refraction index fluid for 193i. One of the failures that just a better knowledge of physics would have prevented.
157i that a nice one !!!!!!
I would really like to see the CaF2 optics getting near water
Even better was the story of the higher refraction index fluid for 193i. One of the failures that just a better knowledge of physics would have prevented.
157i was not supposed to be water, diiodomethane I thinkI would really like to see the CaF2 optics getting near water
Intel invested in 157 nm tools? Maybe the SVGL/ASML Microscan. Pretty much all work on 157 nm stopped around 2004, twenty years ago. Back when Intel was still kicking ass.
There was actually only one 157nm tool I remember delivered: the ASML/SVG/PE Micrascan to imec. In the meantime ASML was working on the immersion 193 prototype and, as soon as the first images printed with it came out, 157nm was dead. Intel invested a lot in programs in 157nm over the years and even placed orders but I do not remember them ever receiving a real tool, maybe one of the small "lamps" from Exitech. Those for me were not a litho tool. As an old friend told me once, "litho is about alining two good images on top of each other, not about printing a pattern".
As for 157i I never heard of it and I am not familiar with the compound. As all attempts by people forgetting the Kramers-Kronig relations at high n fluids failed I just though about water and it was just funny thinking about it and CaF2.
We are talking before 2004 and EUV was studied but it took still years before getting to the first prototype.
As for 157i I never heard of it and I am not familiar with the compound. As all attempts by people forgetting the Kramers-Kronig relations at high n fluids failed I just though about water and it was just funny thinking about it and CaF2.
We are talking before 2004 and EUV was studied but it took still years before getting to the first prototype.