Chinese semiconductor industry

Status
Not open for further replies.
S

supersnoop

Major
Registered Member
FairAndUnbiased said:
90 nm isn't bad at all. It is literally at the same level as Canon, a historical giant in optical technology. SMEE has achieved commercial success in packaging lithography applications (40% market share). They laugh but they don't realize that it already has commercial success and "good" vs "bad" is a far smaller gap than "having" vs "not having".

Even as a front end technology, 90 nm is plenty for even some pretty high end microcontrollers that can do everything except run high fidelity graphics and real time DSP. For instance, programming an embedded chip like ARM or AVR, you realize that you can do quite a bit (as in, build some types of scientific instrumentation and machine tools) with 32 bit microcontrollers on a 180 nm process. The cost of scientific instruments and machine tools is rarely the chips, it's the extremely high quality materials and specialty parts.
Click to expand...
I think another thing to note here, correct me if I'm wrong as I'm not an expert
The jump from 90nm down to 45nm is/was prime Moore's Law territory, so it would only take 2-5 years (historically speaking)
I also believe there are no native Japanese processes running smaller than 45nm anyway, so in theory SMEE is in position to cash in on chips for most industrial chips (which is the bread and butter of Japanese production)
 
FairAndUnbiased

FairAndUnbiased

Brigadier
Registered Member
supersnoop said:
I think another thing to note here, correct me if I'm wrong as I'm not an expert
The jump from 90nm down to 45nm is/was prime Moore's Law territory, so it would only take 2-5 years (historically speaking)
I also believe there are no native Japanese processes running smaller than 45nm anyway, so in theory SMEE is in position to cash in on chips for most industrial chips (which is the bread and butter of Japanese production)
Click to expand...

90 nm to 45 nm was not really Moore's law era which was really 80s and 90s (going from microns to nanometers), it was the start of the slowdown.

The major innovation from 90 nm to 45 nm was not a simple die shrink. They needed a new chemistry for the gate - changing oxide gate to high k gate to avoid dielectric breakdown. They also had to start using ArF double patterning, this was the last node possible for KrF lasers. Some companies began deploying immersion ArF for this too.

Fortunately on the chemistry side Chinese companies are quite well prepared.
 
W

WTAN

Junior Member
Registered Member
tokenanalyst said:
I really don't know. But the posts make it seem like if they are building the machine from the ground up.
Click to expand...

ansy1968 said:
@WTAN Sir the foundation of SSA800 DUVL by adding immersion?
Click to expand...
I believe it was a gradual development over time which led eventually to the 28nm DUVL.

I posted an article a couple of years ago that claimed that SMEE has already sucessfully developed the technologies for a 22nm Node DUVL.
Probably uses Improved Light source and Advanced Optics.

This will probably be released as an improved model specifically optimised for 7nm and even 5nm IC Production. When it will become available depends on the demand from Chinese FABS for a DUVL capable of 7nm IC Production.
SMEE will continue to improve the DUVL till an EUVL becomes available.

For the time being the main task for SMEE is to Serial produce the 28nm DUVL and deliver the units that are already on order.
 
W

WTAN

Junior Member
Registered Member
tokenanalyst said:
To China? I believe that ASML and even the Dutch government wants to continue dominating the Chinese market, they want to continue selling at least their DUVi machines to China. The problem is that the United States has groups that want to restrict or even destroy the Chinese semiconductor industry because it is a strategic industry. ASML and to some extent Nikon are key to that goal. You can have all semiconductor manufacturing equipment and software, but if you can't print the patterns of your chips you can't make them. For now the goal of the American government is to set China at 14-10nm, but if Trump wins and the Republicans take power, the goal could become to destroy the entire Chinese semiconductor industry. And I say this without hyperbola, I believe China should make its semiconductor industry independent of the United States as much as possible. Americans normally bully countries that are highly dependent on foreign technology. If the Chinese achieve a certain level of technological independence, the Americans are going to back down because their companies are going to complain.

Please, Log in or Register to view URLs content!
Please, Log in or Register to view URLs content!
Please, Log in or Register to view URLs content!
Please, Log in or Register to view URLs content!
Click to expand...
IN the YAHOO Article, the Dutch acknowledge that the ASML Monopoly on EUVL Machines will last up till 2025.
After that they will lose the EUVL Monopoly to a competitor.
Thats an acknowledgment that the Chinese EUVL will be available sometime in 2025.
 
Status
Not open for further replies.
Top