Interesting and detailed article on this key technology for lithography machine.
Few takeaways:
1. US monopoly of Keysight (formerly Agilent) and Zygo has been broken.
2. Harbin Institute of Technology built a total of 10 interferometers, for SMEE and some other institutions
3. Confirmation that exists at least one SMEE prototype for 28nm ("lithography machine prototypes at multiple process nodes from 350 nm to 28 nm")
4. Confirmation that SMEE machine is fully localized. Interferometer from Zygo was considered a critical sub-part of the machine, but now it seems they can do without. We already now RSLaser replaced Gigaphoton as light source.
5. They are now working on the interferometer for next-generation EUV lithography machines
Great summary there. I think all this noise from the past few months indicates that the development of SMEE's Arf scanner is coming to a close and they are getting ready for mass production. Given SMEE has recently been put on entity list, it's probably a good mention that they are fully localized.
Interesting that the Laser Interferometer mentioned here has been mentioned by havok as something they use that's sufficient for a scanner with 1950i performance, but they need something better (grating interferometer) for 2nd gen DUVi scanner. So, they are still probably 2 years away from having a domestic option that's as good as NXT2000i/2050i. Hence the need for SMIC to accelerate its purchase of ASML scanners right now.
All these recent unveils point to 2 things:
1) They are getting closer to building a viable prototype for EUV
2) They are releasing the news to put pressure on Dutch gov't and ASML to not take their monopoly position for granted and buckle under US sanction pressures.
According to havok，the domestic production line at ICRD is only starting volume production of 55nm，28nm is still long way to go
View attachment 104141
Yes, Given the recent news about Arf dry 光刻胶 and 涂胶显影机 and lack of comment about immersion for that (that
@tokenanalyst and
@PopularScience posted), they are just in the process of testing domestic production line for 55nm. Photoresist for Arfi still needs some time.
I'm expecting the Arf immersion scanner to be testing with SMIC's Beijing de-americanized line. There was that story of SMIC Beijing having 40nm de-americanized line and then getting to 28nm. Given the likelihood of further sanctions, I think they will have production line testing SMEE scanner for 40 nm production and then 28nm production later.
There is this comment
现在国产DUVi光刻机到底啥进展啊？在测试了吗？
12-29
· IP 属地陕西
1
作者
接近交付，只是交付客户不是量产。
For
@tokenanalyst
不光华卓，启尔，国望等现在都是靠国家补助研发的，没有补助靠企业自身的盈利哪有钱搞这种研发
仔细看了一下，那个招股书多半说的是财务上的问题，02专项是专款专用，和公司其他项目不掺合
In case you are wondering. Everyone is getting state subsidies to do these developments. 02 project related funding/news are not reported for good reason.
So if I was to summarize this. The new SSA/B/C-800 is dual stage and getting plenty of gov't funding. I think it will replace SSA/B/C-600 line and will actually be commercially competitive. the Arf dry version seems to be more capable and able to handle 55nm production (big improvement over SSA-600 which can only do 90 to 65 nm??). They have now developed full domestic process for up to 55nm that's been tested in production environment at ICRD. Krf and iLine products are ironically not finished development yet due to the greater urgency for Arf production line. Which means, they will likely start mass production of Arf dry scanner soon. Sounds correct??
DUVi has been delivered (probably SMEE and maybe YMTC) since they can use non-Chinese tools for their regular process. They have not started mass production with that. Likely, they are still working on their process. Mass production of DUVi is unlikely to start until the customers are comfortable with the machines. That is unlikely to happen until June at least.
@PopularScience where did you see the comment about June for mass production of immersion scanners?
Last year, news came out that SMIC have received the 11 AMSL DUVs for $1.2B. Each one would cost $100M, that indicates its the most advanced version NXT2050, capable of 7nm, but hvpc indicares that ASML wouldn't deliver so many advanced ones in such short time. So , that leads to so what conflcted views. Maybe it could something for your research.
hmm, it was not 11 ASML DUVi for $1.2 billion. They just said they spent that much money without specifying which scanners they bought. It was likely that they bought a few NXT2050i on that order and a lot more of the other scanners. I think they have enough (or will have enough) ASML machines to complete their SN1/SN2 expansion and for some of their Beijing & Lingang lines also.
I'm not an industry insider so maybe they are right BUT as a consumer why should I pay an overprice American Brand when I can buy a cheaper Chinese one. This week my wife gifted me a Huawei Matebook 14 notebook laptop (I'm using it today typing this post
