Chinese semiconductor industry


krautmeister

Junior Member
Registered Member
@krautmeister Sir thanks for your explanation, I learned a lot, from what you posted, China is in the forefront regarding research of new tech especially the power source which is lacking at the moment. Sir from what I read its a collaboration is there a problem regarding IP? the US is using legal means to disrupt Chinese tech development.
No issues with IP. The paper was essentially a research collaboration between China and Germany. Any IP coming from this would be process related, IP which has yet to be created since that IP is coming in the next few years from the new Beijing R&D facility geared towards creating a SSMB EUV lithography light source.
 

krautmeister

Junior Member
Registered Member
A synchrotron light source? That seems kind of impractical because the machine will be huge.
You will need to design the facility around the synchrotron.
I can see it being used for either research or even military or government production but for commercial use seems kind of complex.
Word is that it is not as cumbersome as you might think because it is clearly worded as suitable for industrialization, specifically as a lithography light source. How that is realized, we don't know yet but the current timeline states it should be ready sometime this decade....who knows when that could be.

Imo, the more interesting news is the CETC 35nm DUV machine. If the rumored NA 1.35 lens actually exists, that only leaves the problems with overlay precision to overcome. Multi-patterning down to 7nm like SMIC's N+2 process would be possible with domestic equipment by 2023-24 if true.
 

WTAN

Junior Member
Registered Member
1.35NA Optics has been succesfully developed as part of Special Project 02. The 0.75NA Optics was developed many years ago for the SMEE 90nm DUVL.
According to an SMEE Insider the Protoype 1.35NA Optics has been delivered to SMEE late last year for 28nm DUVL Prototype assembly.
How fast the SMEE DUVL can do 14nm or even 7nm Chips depends on making improvements to many aspects of the Machine, but the main improvement should be in Overlay Performance.
This improvement requires cooperation and feedback between FABS and SMEE.
ASML worked with TSMC and SAMSUNG in making improvements to its DUVL range.
 

Quickie

Major
If it has a narrower frequency band, that should mean its more reliable even with less mechanical controls? What's the standard deviation of the 13.5nm LPP light source compared to the current light source?

The way I look at it, the narrower the frequency band of the EUV source, the smaller is the range of deviation in the angle of refraction/reflection.

This way, a sharper image can be achieved with the same mirror precision when compared to the case when a less narrow frequency band is used.
 

Oldschool

Junior Member
Registered Member
Yes......i have read this news about the sucessful development by CETC of the 35nm DUVL.
This is a Major Breakthrough by CETC as this means that in term of Lithography Technology it has caught up with SMEE.
The CETC 35nm DUVL is basically equivalent to the SMEE 28nm DUVL.
SMEE and CETC uses different performance indicators for the DUVL.
CETC folllows ASML and describes the performance of the machine using Resolution. ASML 2000i/1980i Resolution is 38nm. CETC new DUVL resolution is 35nm. So CETCs new DUVL is equivalent to ASML 1980i.
SMEE chooses to describe its DUVL as having a 28nm Node Performance. Node means it can produce 28nm Chips using a Single Exposure.
But basically the New CETC Machine has the same Performance as the SMEE 28nm DUVL.
This is the case as it uses the same Subsystems and Parts as the SMEE DUVL.

This is a extremely significant event as China now has 2 Manufacturers of DUVL Machines.
CETC being the larger Manufacturer with larger scale and talent as well as finacial resources will now compete with SMEE.
CETC will also now certainly play a part in the development of China's first EUVL.
Hopefully cetc can be more aggressive and its 35nm NA1.35 can go to 14nm as well like asml

smee which makes 28nm only and plans to deliver maybe 3 samples this year and mass production begins 2022 and 2023. 14nm won't comes out till 2024.This is according to the big shrimp havok

Dry dual stages are more ready, goes to 65nm . Production is currently.

I wonder how it going with Huawei 40nm which uses that machine
 
Last edited:

Oldschool

Junior Member
Registered Member
No issues with IP. The paper was essentially a research collaboration between China and Germany. Any IP coming from this would be process related, IP which has yet to be created since that IP is coming in the next few years from the new Beijing R&D facility geared towards creating a SSMB EUV lithography light source.
There is already a commercial company doing LPP in Shanghai. SSMB still at academic phase.
 

jfcarli

New Member
Registered Member
Let me try to understand something: China consumes more than 50 pct of semiconductors manufactured in the world. It is, therefore, the most important client of any foundry.

China is now making a huge effort to become self sufficient and is pouring billions and billions of dollars for that matter.

It makes sense for China to invest such humongous amounts of cash in the industry, after all it will be simply replacing imports.

The US is now also alledgedly pouring billions of dollars in order to "onshore"production of chips.

Questions:
a) who are the US going to sell that additional production to?
b) who is TSMC going to sell chips to?
c) why is TSMC still investing in fixed assets outside of China?
 

vincent

Senior Member
Let me try to understand something: China consumes more than 50 pct of semiconductors manufactured in the world. It is, therefore, the most important client of any foundry.

China is now making a huge effort to become self sufficient and is pouring billions and billions of dollars for that matter.

It makes sense for China to invest such humongous amounts of cash in the industry, after all it will be simply replacing imports.

The US is now also alledgedly pouring billions of dollars in order to "onshore"production of chips.

Questions:
a) who are the US going to sell that additional production to?
b) who is TSMC going to sell chips to?
c) why is TSMC still investing in fixed assets outside of China?
Simple, the US will simply ban all products that contain Chinese made chips
 

jfcarli

New Member
Registered Member
Simple, the US will simply ban all products that contain Chinese made chips
What about the rest of the world? If you add what will be used in and by China, with what will be used by The European Union and the rest of the world, I don't think there is much left for the US to consume and therefore sustain the billions it is investing.

One must not forget that a considerable amount of electronics is manufactured or at least assembled in China.

Something does not seem to add up here!
 

Top