Chinese semiconductor industry

Status
Not open for further replies.

tokenanalyst

Brigadier
Registered Member
Yes thats the main reason the US is getting nervous and one of the big reasons SMIC is banned by the US
Not exactly banned but export controlled and looks like they are getting nervous about the possibility of SMIC getting alternative equipment and supply from either locally or from other countries, so there is a complicated war of words between the national security hawks-the defense department and the Commerce department-the U.S semiconductor industry.
 
I dont want rumor. I want real progress with evidence. Where is N+1,2,3,4,5 they talked about since 2 years ago. Atleast SMIC should do 12nm for small progress instead stuck and stay 14nm like forever. Forget about 7nm or euv just do atleast 12 nm.

Catching up in the advanced processes is not the primary goal. Having a fully indigenous supply chain is far more important. Its a much bigger win to do 14nm using fully indigenous processes and equipment than to do 7nm.

Intel was stuck at 14nm for a decade. You only go for the 12nm and 10nm processes when you are like Intel and truly hit a break wall at 14nm. Otherwise you would go direct from 14nm to 7nm.

With 14nm chips, you can do just about everything except for top end mobile phones. Heck, the best desktop CPUs you can get today from Intel are still 14nm.
 

tokenanalyst

Brigadier
Registered Member
I dont want rumor. I want real progress with evidence. Where is N+1,2,3,4,5 they talked about since 2 years ago. Atleast SMIC should do 12nm for small progress instead stuck and stay 14nm like forever. Forget about 7nm or euv just do atleast 12 nm.

You think making ICs is like making bricks? Is REALLY HARD and as node get smaller it will get even harder and expensive, the fact the SMIC was able to make it to 14nm in this race is outstanding. We maybe almost reaching the economic limits and practical limits of making smaller transistors, that is why advance packaging has become such good alternative for scaling.

1646691401962.png
 

ZeEa5KPul

Colonel
Registered Member
Catching up in the advanced processes is not the primary goal. Having a fully indigenous supply chain is far more important. Its a much bigger win to do 14nm using fully indigenous processes and equipment than to do 7nm.

Intel was stuck at 14nm for a decade. You only go for the 12nm and 10nm processes when you are like Intel and truly hit a break wall at 14nm. Otherwise you would go direct from 14nm to 7nm.

With 14nm chips, you can do just about everything except for top end mobile phones. Heck, the best desktop CPUs you can get today from Intel are still 14nm.
China has to eventually get to fully indigenous < 7nm nodes, and the sooner the better. It's true that while today these nodes are almost exclusively for flagship mobile phones, they'll proliferate to other applications eventually. China should be ready to at least compete in these domains by then.

China's position is good-ish today, but that's not a call for complacency.
 

tokenanalyst

Brigadier
Registered Member
I think Chinese companies should focus in creating a 100% indigenous, controllable, secure ecosystem of software, SME and materials around mature nodes 28nm+ and still collaborate with foreign companies for the leading edge, that will allow the China electronics industry to start from a very high point instead from zero in case the U.S China-hawks decide to sacrifice U.S semiconductor companies and jobs for sake of "National Security", while at the same time keeping up with the leading edge.
 

antiterror13

Brigadier
Catching up in the advanced processes is not the primary goal. Having a fully indigenous supply chain is far more important. Its a much bigger win to do 14nm using fully indigenous processes and equipment than to do 7nm.

Intel was stuck at 14nm for a decade. You only go for the 12nm and 10nm processes when you are like Intel and truly hit a break wall at 14nm. Otherwise you would go direct from 14nm to 7nm.

With 14nm chips, you can do just about everything except for top end mobile phones. Heck, the best desktop CPUs you can get today from Intel are still 14nm.

Correct, only just recently (a month or so ago) the best Intel desktop CPU 12th Generation moved to 10nm. before that all were just 14nm for ~10 years .. infamous for 14++++++++++
 

tokenanalyst

Brigadier
Registered Member

Guowang Optical Lithography Machine Exposure System Production Base Project has made new progress, and three companies are listed as winning candidates.​


The Beijing Public Resource Trading Service Platform recently released the “Announcement of Winning Candidates for the Clean Project of Projection Lithography Exposure Optical System R&D and Mass Production Base Project”, Silian Intelligent Technology Co., Ltd., etc.Three companies are among the winning candidates.

According to public information, the construction unit of the production base is Beijing Guowang Optical Technology Co., Ltd., and the core R&D team members are from Changchun Guoke Precision, the only ultra-precision optical system industrialization technology research and development unit in China. The basic positioning of Guowang Optics is to establish my country’s independent ultra-precision optical industry technology research and development and production system, and to form the mass production and supply capacity of extremely large-scale IC manufacturing projection lithography machine exposure optical systems, high-end precision optical instruments and equipment, fundamentally Solve the industrialization problem of my country’s IC manufacturing projection lithography machine exposure optical system and high-end precision optical equipment.

Guowang Optical R&D and production base is located in Beijing Economic and Technological Development Zone, with a planned construction area of 120,000 square meters. it will have the R&D, design and mass production and supply capabilities of 350/280nm node, 90/110nm node, 28nm and below nodes for extremely large-scale IC manufacturing projection lithography exposure optical system products.
 

xypher

Senior Member
Registered Member
That's still not too bad. SMIC is getting one hand tied behind their back in this and being able to keep up in finFET at all is incredible considering many giants have dropped out of that race.
Agree. There are basically only 3 other companies that can produce at 7 or 8 nm - Samsung, TSMC & Intel. Plus I think I read somewhere that SMIC has finished the R&D for 3\5 nm nodes and is now waiting for the opportunity to develop & iron them out. So had it not been for the tech embargo, SMIC could have been the 3rd company to reach 3/5 nm nodes.
 

european_guy

Junior Member
Registered Member
I think Chinese companies should focus in creating a 100% indigenous, controllable, secure ecosystem of software, SME and materials around mature nodes 28nm+ and still collaborate with foreign companies for the leading edge, that will allow the China electronics industry to start from a very high point instead from zero in case the U.S China-hawks decide to sacrifice U.S semiconductor companies and jobs for sake of "National Security", while at the same time keeping up with the leading edge.

I think China maybe should fix a point target in the future and work backward.

In November 5, 2024 there will be the elections in US, there is a high probability (according to current polls) that some very hawkish republican will be the new president. So maybe starting 2025 there will be very possibly a strong ban on China high-tech. With that target in mind:

1) In which (realistically) state of local semiconductor ecosystem China wants to be at that time target of early 2025?

2) Which step are needed? Which is the plan to reach the target at point 1 starting from current situation?


P.S: Eventually a ban could be enforced also earlier if the election campaign gets tough and current administration needs some "tough on China" show off.
 
Status
Not open for further replies.
Top