Chinese semiconductor industry

Status
Not open for further replies.

WTAN

Junior Member
Registered Member
@huemens @WTAN @foofy @krautmeister if the USG do blacklist Honor, will this trigger the Chinese gov't to retaliate and order SMIC to supply chips to them and Huawei. Since the 14nm domestic line will be introduced next year will the Chinese risk it? I know the CPC are rational thinker but this is the time to strike back and force the issue with chip shortage lasting until 2023. The trust has long been broken and by nitpicking every successful Chinese tech company, this ain't about national security concern it is an economic warfare.
My take on this is that the US GOVT may not do anything.
I believe that there is already a concensus by Major US Think Tanks and Researchers that the Semiconductor Sanctions on Chinese companies are counter productive and will cause the Chinese to set up their own independant production lines.

The people still calling for sanctions to be imposed on YMTC and Honor are a bunch Republicans led by Marco Rubio which are still pushing for Trump style sanctions.
So far we havent seen any new sanctions on YMTC.

Sanctioning Honor might damage its business but it will push many more Chinese Mobile Manufacturers to decide to use Harmony OS and the resulting demand for locally made ICs will only accelerate the develoment of local IC production capability.

Yes.....the Chinese Govt could retaliate by forcing
TSMC Factories in China and SMIC to supply Honor under the new Entity List Laws.
 

AndrewS

Brigadier
Registered Member
China CURRENTLY isn't in the position to retaliate. The power discrepancy in semiconductors between China and the American semiconductor alliance is just too great. Even if DUVi 14nm dual patterning was here now in 2021 and yielding economically, it would still be in a position of weakness because of the lack of scale. Until it can supply sanctioned items to a critical level, let's say 50% of domestic IC consumption, any retaliation will invite an immediate escalation where America CURRENTLY has relative strength vs China, despite the chip shortage. Note that I randomly said 50% of domestic IC consumption NOT domestic consumption PLUS ICs for integration into exported products. Once such a threshold is passed will the pain start becoming greater for America than China. We don't know what the threshold is exactly, but I know for sure we haven't reached that threshold yet. Huawei is a far more important entity than Honor so if China isn't willing to retaliate on behalf of Huawei, then Honor would be sacrificed for sure.

The bottom line is, as long as China can achieve technological capability to produce 14nm economically, it's all over for the American led semiconductor alliance. In this scenario, quad-patterned 7nm will be almost guaranteed to follow. When that happens <=5nm EUV chip margins will collapse, which will allow China the time it needs to catch up completely. When that time comes, the members of the American led anti-China semiconductor alliance will have a tough choice. Either they break ranks and become a tech leader of the newly formed China semiconductor supply chain while retaining the China market or they stick with the anti-China semiconductor alliance and become a price uncompetitive piece of a shrinking supply chain with fewer profits for R&D relying on government subsidies and probably eventual obsolescence. The anti-China semiconductor alliance will go down in history as the reason why China became a predominant semiconductor power.

Agreed.

Adding 50% of current global semiconductor capacity inside China would need a huge number of new fabs in China.

And now it has become a national security issue with the technology war, I fully expect China to mobilise and build these fabs as soon as the technology is ready. We've seen this happen in many other cases both in the military and commercial realms.

A conservative methodology is below.


In December 2020, global monthly capacity of semiconductors was 21 million wafers per month (200mm equivalent)

So if you wanted to add an additional 50% of current global capacity inside China, you could need the following combination:

38x Huge Fabs (120,000 wafers per month @ 300mm)
AND
131x Large Fabs (80,000 wafers per month @ 200mm)

Once domestic Chinese technology is mature, China could start mass-producing semiconductor equipment.
I imagine it would be like producing dumplings / sausages.

Due to economies of scale and lower costs, theoretically Chinese semiconductors would be the LOWEST cost in the world.

And yes, a big problem would be finding customers for this capacity.
But also ramping up expertise in operating the fabs and designing chips.

But remember that memory chips are commodities and have the biggest demand. Plus they are most cost effective at older process nodes which use DUV lithography.
 

caudaceus

Senior Member
Registered Member
This article is not accurate.
This article was already discussed a month ago.
All components have already been delivered to SMEE for assembly.
The SMEE 28NM DUVL should be delivered to end users later this year.
From that article, light source is the major problem. Is it been solved?
 

Vichysoy

New Member
Registered Member
According to the SMEE Insider, Light Source has already been shipped to SMEE for integration.
Problem with the Light Source has already been solved.
I gave some details about the Light Source several weeks ago in a similar discussion with another member.
Thanks bro ! But who solved the light source ?
 

Xizor

Captain
Registered Member
Lenovo may develop chips in the future

Yang Yuanqing, chairman of Lenovo Group, said in an interview that Lenovo Group’s R&D investment increased by 40% last quarter, and strives to double in the next three years. The company will focus on investments in new IT fields, as well as components such as display, battery, chips and even software. At the same time, Lenovo may develop its own chips.
 

FairAndUnbiased

Brigadier
Registered Member
Lenovo may develop chips in the future

Yang Yuanqing, chairman of Lenovo Group, said in an interview that Lenovo Group’s R&D investment increased by 40% last quarter, and strives to double in the next three years. The company will focus on investments in new IT fields, as well as components such as display, battery, chips and even software. At the same time, Lenovo may develop its own chips.
Good on Lenovo but it'll probably only be peripheral chips that have sufficient value added like display drivers and battery management for now. Developing an x86 CPU from scratch will be difficult and expensive.
 
Status
Not open for further replies.
Top