Chinese semiconductor thread II

PopularScience

Junior Member
Registered Member
All 3 presenters in this video provided valuable info regarding Chinese EUV development. I’m a little surprised such info was allowed to be publicly available. Regarding high power high repetition CO2 laser, here is prof Tang confirming in 2023 that HUST CO2 laser main amp achieved 10kw output, and seed/preamp lasers were going very well. Hopefully they can get to at least 20kw now, which should be sufficient for risk maybe hvm production assuming reasonable CE
Please, Log in or Register to view URLs content!

from internet

NXE: 3400B is ASML's first generation of EUV that is truly put into mass production. Its LPP power is 250W and its laser power is 20kW. The latest news is that the domestic level in the middle of last year was several kW, which seems to be far behind. However, in fact, ASML's 20kW laser power is composed of more than one 13kW laser connected in series, so the gap is not so big at first glance.
At the same time, NXE: 3400B is not ASML's first generation of EUV. There are several generations before it, but they are only experimental equipment and have not been put into mass production lines. The previous generation of 3400B is 3300B, whose lpp power is only 80w. This model itself has been mass-produced and sold to some research institutions for testing. In other words, the 80w 3300B can also produce chips, but the cost is high and the efficiency is low, which cannot meet the requirements of TSMC and Samsung's mass production lines, so it can only be sold to some chip research institutions.
In this way, even if the laser stops at a few KW forever, it is no problem to build a 3300B-level EUV now. Such equipment can also be put into mass production lines as a last resort, and it still produces 13.5nm extreme ultraviolet light, and can still produce 5nm3nm, but the cost is higher (Isn't Kirin 9000s1 such a high-cost product as a last resort?)
 

sunnymaxi

Major
Registered Member
from internet

NXE: 3400B is ASML's first generation of EUV that is truly put into mass production. Its LPP power is 250W and its laser power is 20kW. The latest news is that the domestic level in the middle of last year was several kW, which seems to be far behind. However, in fact, ASML's 20kW laser power is composed of more than one 13kW laser connected in series, so the gap is not so big at first glance.
At the same time, NXE: 3400B is not ASML's first generation of EUV. There are several generations before it, but they are only experimental equipment and have not been put into mass production lines. The previous generation of 3400B is 3300B, whose lpp power is only 80w. This model itself has been mass-produced and sold to some research institutions for testing. In other words, the 80w 3300B can also produce chips, but the cost is high and the efficiency is low, which cannot meet the requirements of TSMC and Samsung's mass production lines, so it can only be sold to some chip research institutions.
In this way, even if the laser stops at a few KW forever, it is no problem to build a 3300B-level EUV now. Such equipment can also be put into mass production lines as a last resort, and it still produces 13.5nm extreme ultraviolet light, and can still produce 5nm3nm, but the cost is higher (Isn't Kirin 9000s1 such a high-cost product as a last resort?)
you are senior member on this forum and knowledgeable too. even you know that, the Chinese internet is not the accurate source.

20KW laser output is likely to achieved by HUST now.. they have had working 10KW laser in mid 2023. officially announced by the institute and now Lithography seminar 2024. so this ''only several KW'' information is outdated.

CIOMP/Harbin and SIOMP all achieved sufficient laser output for first generation Chinese EUV machine.

actual laser patent.png
CIOMP, LPP-EUV laser patent.. announced in 2024 but filled in 2022.

HIT_EUVLightSource-2024EOY.jpg

HIT delivered 13.5nm LPP-EUV light source for integration.. this happened months ago. they have won provincial prize as well.

Link -
Please, Log in or Register to view URLs content!
 

PopularScience

Junior Member
Registered Member
you are senior member on this forum and knowledgeable too. even you know that, the Chinese internet is not the accurate source.

20KW laser output is likely to achieved by HUST now.. they have had working 10KW laser in mid 2023. officially announced by the institute and now Lithography seminar 2024. so this ''only several KW'' information is outdated.

CIOMP/Harbin and SIOMP all achieved sufficient laser output for first generation Chinese EUV machine.

View attachment 142463
CIOMP, LPP-EUV laser patent.. announced in 2024 but filled in 2022.

View attachment 142464

HIT delivered 13.5nm LPP-EUV light source for integration.. this happened months ago. they have won provincial prize as well.

Link -
Please, Log in or Register to view URLs content!
I just want to point out maybe they can chain several CO2 lasers together to produce stronger output, like the first generation ASML EUV.
 

broadsword

Brigadier
HIT delivered 13.5nm LPP-EUV light source for integration.. this happened months ago. they have won provincial prize as well.

Link -
Please, Log in or Register to view URLs content!

The breakthrough must be significant and most likely at the commercialization level for it to win the prize. No way it could have won if it was good only inside the lab.
 

sunnymaxi

Major
Registered Member
The breakthrough must be significant and most likely at the commercialization level for it to win the prize. No way it could have won if it was good only inside the lab.
absolutely. this is basically official confirmation.

There is no way to know if a tech is commercial level or not until you actually produce. Something may sound good on paper but commercializing is a much different story.
this is not on paper. if you follow their development process. they have been working on this for years. this is likely the first prototype of 13.5nm LPP-EUV light source delivered for integration into a full machine.

we have other source as well. integration of first prototype is happening.
 

latenlazy

Brigadier
That doesn't sound good.

Japan's Gigaphton CO2 laser for EUV output power is 20KW. That's 6 times of this.

I think this HUST laser is lab demo purpose , not really for commercial production purpose.

I think this University program needs to partner with state owned laser weapon companies in order to bump to that level of power. Several room sized equipment is beyond university program calibur.

Please, Log in or Register to view URLs content!
A seed laser is not the main laser. You are comparing an apple seed to an apple.
 

latenlazy

Brigadier
Prof Tang’s team is responsible for RF slab CO2 laser, which is used in the preamp step of CO2 MOPA system. Trumpf’s preamp power is 100 W, so preamp power is not a problem. That 3.5KW is probably for spike laser annealing applications

Please, Log in or Register to view URLs content!


Another team at HUST is responsible for mopa main amp laser development. They achieved 10 KW level in 2023, so power is no big issue

Please, Log in or Register to view URLs content!


Also don’t forget dose control and beam quality parameters are as important as power for lppeuv drive laser systems.
That doesn't sound good.

Japan's Gigaphton CO2 laser for EUV output power is 20KW. That's 6 times of this.

I think this HUST laser is lab demo purpose , not really for commercial production purpose.

I think this University program needs to partner with state owned laser weapon companies in order to bump to that level of power. Several room sized equipment is beyond university program calibur.

Please, Log in or Register to view URLs content!

Just to be clear, the Chinese doesn’t say “10 kilowatt” as in the specific number but “ten(s of) thousand watt” as in the general order of magnitude.
 
Top