Chinese semiconductor industry

Status
Not open for further replies.

latenlazy

Brigadier
Why is there complete radio silence on this issue now? I believe that there is a domestic program to build a 28nm DUV machine and I believe it will succeed. But i doubt the very optimistic timetable that they can deliver on it last year or even this year.
Because not every technological breakthrough is going to get a giant media reception? Especially if they’re geopolitically sensitive?
 

olalavn

Senior Member
Registered Member
Havok is just a poster on a forum not different from us. He doesn't post any links to back up his claims. There are no official announcements on this delivery either from the government, media, SMIC or for SMEE self. Unless I see some official announcements I remain skeptical about the claims that the SMEE 28NM DUV machine is ready now.
I only doubt how CETC and SMEE will compete in Mainland... ASML's DUV-i is still huge... and it's too politically sensitive...
 

Franklin

Captain
Because not every technological breakthrough is going to get a giant media reception? Especially if they’re geopolitically sensitive?
China actually benefits geo-politically from making its progress on lithography technology public as that will make other countries think twice about banning sales to China.
 

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
The way they used to validate 90 nm scanner was a totally failure,SMEE didn't have commercially viable production in mind when they develop the machine. It was merely a demo project designed to be pass verification by government officials,with no practical industrial use case. The machines eventually end up in dusts in domestic fabs because they were useless.
Well, that's what happens when domestic foundries are not working with them.

Keep in mind that certification happened in 2018 and that it was only reported to be in production by Shanghai gov't. As such, it likely took 4 years to reach a point where it's commercially viable. So I presume that at least one fab (Yandong at least) is using it commercially. Now, that doesn't mean it's actually competitive to foreign scanners. Hence, the new line of SSA/B/C-800 dual stage scanners.
Now they have learned the lesson and changed the way to validate these machines,the machine need to uphold the test on real production line and the performance need to be acceptable by fabs in order to pass the validation.

Yes,SMEE is behind schedule. Arf delayed by 3 years,Arfi delayed by 2 years
Again, this project never should've been solely left with SMEE to work on. SSA600 was probably developed to be on par with whatever CETC uses. Which means it works, but likely has much lower throughput and such. That kind of product works if you have nothing else, but can't compete with any real competition. If this project was developed with closer integration with fabs, they would've been told that it needs to meet this level of throughput for us to be able to use it commercially. Looks like they are finally able to do that.

My sense is Havok only started to post more frequently because he sees an end to the development. Otherwise, it would be too embarrassing to keep telling everyone that the product you've been working on is delayed for another year.

And based on his comment about expanding production + new production line, I think we should feel very confident about commercial success of SSA800 dry version next year. Someone from SMEE is not going to say that unless they have a product now that domestic fabs are actually going to order.
I read somewhere that SMEE 90nm used many foreign components, including Gigaphoton as light source.

We don't know why it was dead on arrival, we can assume prototype didn't work and didn't reach targeted specs. But anyhow in the past SMEE used many foreign suppliers. Their new machines are 100% localized (and now they are also on entity list, just to be sure they don't change their mind).
Right, it seemed like SSA600 originally was using foreign components from critical components. It looks like SSA800 is using domestic stuff for the really important components. Hence the importance in the light source, laser interferometer, lens, immersion system being domestic that I think he discussed.
My impression is that in the past they were left more or less alone in their development, yes they were granted funds, but that's all. Today they and their key suppliers have strong engineering support from Huawei, ICRD and many national research institutes. China realized (a bit late IMO) that lithography machine development requires a nation level effort.
Unfortunately, it took sanctioning Huawei for them to realize it.

Apparently,the new 28nm prototype machine contains many foreign parts. It takes time to replace all of them,according to havok


View attachment 104227
where are you seeing this comment? I looked through it several times and don't see it. Having some foreign parts is not an issue as long as they are stocked up and can be sourced. Frankly, just posting his comment without context doesn't help.

He seems pretty clear that the issue was product not being commercially competitive rather than usage of foreign parts. Which tells me this is not an issue.

He is referring to the immersion 28nm machine of SMEE. Previously stated that it was ready to be delivered....it has now been delivered to customer for validation and acceptance testing.

This is a milestone. A good way to start 2023.
So based on the struggles of SSA600 to be employed by fabs (took 4 years from certification to be reported as in production by Shanghai gov't), I had always assumed you'd need at 3 years from 2021 certification to be put into usage by fabs. Seems like they are basically on track for that. It's likely they will have a couple of prototypes delivered to SMIC, ICRD, Huawei and YMTC which have already passed initial certification. Now, they have been delivered to customers but customers are probably still getting things set up and haven't started production yet. This is probably going to take a year to get all the issues sorted out.

The fact that he showed photo of ICRD line for Arf dry tells me that ICRD already has an operational line and it is far enough along in production process that they are close to being mass produced.

I would guess this is a relatively recent event, since his appearance on zhihu only picked up recently. Again, unlikely someone will come on public forums if the project they work on is still delayed.
Why is there complete radio silence on this issue now? I believe there is a domestic program to build a 28nm DUV machine and I believe it will succeed. But i doubt the very optimistic timetable that they can deliver on it last year or even this year.
is it really optimistic? They are 2/3 years behind. The original expectation was that they'd have something in mass production in 2022, but it looks like this will have to wait until 2024.
 

gelgoog

Lieutenant General
Registered Member
The original expectation was that they'd have something in mass production in 2022, but it looks like this will have to wait until 2024.
We already could infer this from the reports about the construction of U-Precision's mass production facilities. But some of you guys were overly optimistic. Without U-Precision's dual stage system how could SMEE mass produce the SSA800?
 

mmbro

New Member
Registered Member
China actually benefits geo-politically from making its progress on lithography technology public as that will make other countries think twice about banning sales to China.
With all the latest semicon development appearing on the first days of 2023, thus it signals the shift of science centers to China/Asia

"The center of world science, for instance, was in France in 1740, before it moved to Germany, then Britain, and, later, America, carrying with it, in each case, a major dimension of global leadership."

Please, Log in or Register to view URLs content!
 

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
We already could infer this from the reports about the construction of U-Precision's mass production facilities. But some of you guys were overly optimistic. Without U-Precision's dual stage system how could SMEE mass produce the SSA800?
Well, there was some expectations that they have existing facilities that can produce for low rate production. Also, going by statement about grating inferometer still being worked on, it sounds like smee has been slower in integrating parts than expected.
 

sunnymaxi

Major
Registered Member
Why is there complete radio silence on this issue now? I believe there is a domestic program to build a 28nm DUV machine and I believe it will succeed. But i doubt the very optimistic timetable that they can deliver on it last year or even this year.
you have been on forum for years. you should have known this.

this is the problem of west. Chinese work exactly opposite how west works. they don't brag , don't do chest thumping.

you will never get an official announcement from SMEE. i can bet on that. coz Chinese mindset is totally different.

DUVi 28nm is done and dusted. by the end of this year or early next year serial production will begin. this is 200% confirmed news.

you said,

''I believe there is a domestic program to build a 28nm DUV machine and I believe it will succeed''

a typical statement. Chinese even crack EUV technologies. let alone DUV. @tokenanalyst regularly post updates/info/patents of EUV subsystems.

who is havok. LOOOL

let me tell you one thing, a local Chinese tech enthusiast have more inside information that you could ever imagine.
 

PopularScience

Junior Member
Registered Member
you have been on forum for years. you should have known this.

this is the problem of west. Chinese work exactly opposite how west works. they don't brag , don't do chest thumping.

you will never get an official announcement from SMEE. i can bet on that. coz Chinese mindset is totally different.

DUVi 28nm is done and dusted. by the end of this year or early next year serial production will begin. this is 200% confirmed news.

you said,

''I believe there is a domestic program to build a 28nm DUV machine and I believe it will succeed''

a typical statement. Chinese even crack EUV technologies. let alone DUV. @tokenanalyst regularly post updates/info/patents of EUV subsystems.

who is havok. LOOOL

let me tell you one thing, a local Chinese tech enthusiast have more inside information that you could ever imagine.
Havok working in SMEE. That’s why he posted insider photos.
 
Status
Not open for further replies.
Top