You and me too! Very excited! Jumping from 28nm to 14nm is from planar to FINFET transistor design. The 14nm to 7nm is much easier by comparison. They might even have a pilot 7nm line end of 2022 at this rate, but that's hopeful thinking on my part.
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
@krautmeister Thanks bro appreciated, I don't want to sound too optimistic having been burned several times but I share the same feeling, so bro a question what comes next? I want to know your view and opinion. One of them is the SMEE 28NM DUVL, can they able to improved it comparable to the latest ASML NXT 2050i DUVL? and if it does then 5nm is a possibility in 2024?
Last edited:
Well brother, this opinion I am pulling out of my vaccinated ass, this is what I think, even though I know nothing about it.
Obviously, there are a lot of moving parts, as the expression goes.
Every part is expensive, and required advanced technical knowledge.
To make the entire system work at maximum capacity and efficiency, will require a lot of effort. It will need capital and it will need the know-how.
This process of fine tuning the complicated system will take time.
However, China still has a 80 hour a week work culture for important stuff, so a big effort will be made.
A key factor here has always been China's size, ever since ancient times. The scale makes it easier. They could have two or three teams working on the same problems or similar programs. Such as get the 14nm line up and running smoothly as quick as possible.
Also, with the labour and STEM graduates, one big project is to domestisize the chip production line, from 28nm to 14nm to 7nm and smaller. The other big product is to work on advanced materials to move beyond Moore's law as it applied to silicone.
Once they are ready, production will run, and the product ie the chip will be sold to another Chinese company for their value added manufacturing.
If some other company gets cut off during these iteration cycles ... what can we say ... that's life!
Butt hurt is a part of life. My ass still hurts after I got vaccinated.
[rant] [/rant]
I just felt like ranting today.
[rant] [/rant]
Hopefully they are keeping this 5nm as a national secret. The EUV also is a secret, and will be ramped up like they ramped up vaccination rates.
@horse bro it had become a three prong approached, 1) SMIC with ASML DUVL 2) Domestic equipment with SMEE DUVL 3) researching new material and an EUVL. Once the domesticated 7nm line is in used those sanction against SMIC will be lifted therefore adding additional capacity making China more competitive. So its a WIN WIN for China and a major headache for TW and SK, instead of relying on them, China had become a major competitor. Life sure is tough being an American bitch...lolWell brother, this opinion I am pulling out of my vaccinated ass, this is what I think, even though I know nothing about it.
Obviously, there are a lot of moving parts, as the expression goes.
Every part is expensive, and required advanced technical knowledge.
To make the entire system work at maximum capacity and efficiency, will require a lot of effort. It will need capital and it will need the know-how.
This process of fine tuning the complicated system will take time.
However, China still has a 80 hour a week work culture for important stuff, so a big effort will be made.
A key factor here has always been China's size, ever since ancient times. The scale makes it easier. They could have two or three teams working on the same problems or similar programs. Such as get the 14nm line up and running smoothly as quick as possible.
Also, with the labour and STEM graduates, one big project is to domestisize the chip production line, from 28nm to 14nm to 7nm and smaller. The other big product is to work on advanced materials to move beyond Moore's law as it applied to silicone.
Once they are ready, production will run, and the product ie the chip will be sold to another Chinese company for their value added manufacturing.
If some other company gets cut off during these iteration cycles ... what can we say ... that's life!
Butt hurt is a part of life. My ass still hurts after I got vaccinated.
[rant] [/rant]
I just felt like ranting today.
[rant] [/rant]
The development is swift I believe because they already have the downstream capability. This reminded me of an opinion I read on Twitter, a consequence of a country building its indigenous MICs, from design, development to acquisition is a better understanding of tactic, strategy and doctrine since the nation as a whole has a deeper understanding of the said military hardware. I think it can be reflected on this.
I'm still shocked they had an actual official say what Wen Xiaojun said. China almost never declares anything publicly like this until they are already almost near completing what they are talking about or if they have cleared over the most difficult tasks. So, for them to say MASS PRODUCTION end of 2022 is almost unbelievable. I was expecting the 14nm trials at the end of 2021 to last almost a year or at the most optimistic, 6 months. Then they'd have some commercial prototype released to SMIC, Huahong, etc. where they would be working on taping out designs at those nodes for at least 5-9 months. That would have brought us well into 2023 before even attempting mass production. This is all assuming everything went without a hitch. Wen Xiaojun is the director of the Electronic Information Research Institute of China Electronics Information Industry Development Research Institute and he said mass production is happening end of 2022. That in China official language usually means, BEFORE end of 2022. If this happens, then there are things nobody on this forum knows about or heard about. My feeling is, all exponentially increased Asian hate crimes over the last 2 years drove a lot of critically essential semiconductor equipment experts back to China because this rate of progress just makes no sense.
The 1st generation DUV immersion lithograph is definitely going to be inferior to ASML machines, mostly because of the light source being underpowered. From what we know, the 1st commercial machine will have a 1.35NA lens for the 193nm ArF light source. The only "major" difference besides the overlay accuracy is whether they will have an adequately powerful light source. Last I heard, the candidate for the light source is the RS222 at 20W with the late stage development model RS244 at 40W and a 60W prototype. With immersion, you really need 60W to get the wafer throughput competitively high enough for mass production or else you'll be at a large cost disadvantage against Nikon and ASML machines. To really ramp up throughput competitive with ASML, you need a light source closer to 90W. All the other things are engineering improvements that are less significant. I'm not hoping for it to surpass ASML DUV machines anytime soon. What I am expecting is for it to break the sanctions so that China's leading tech companies can survive the next 6-8 years when they can finally compete head, with 100% domestic supply chain, with the likes of ASML, TSMC, etc.
I forgot that this is China and their DUV immersion lithographs will probably sell for 3 times less than ASML or Nikon.
Definitely will be competitive, underpowered light source or not.@krautmeister bro a stupid question can DUVL use a DPP light source, Harbin Institute of technology develop a 150W DPP can SMEE used it on their DUVL?I forgot that this is China and their DUV immersion lithographs will probably sell for 3 times less than ASML or Nikon.
I think you're talking about DPP as opposed to LPP light sources for EUV. I am going to take a guess you read that article where the journalist was skeptical of China's semiconductor progress where he compared the ArF light source power to LPP right? That guy mixed up the light sources for DUV and EUV and did a direct power comparison, which is obviously wrong.
- Status