Chinese semiconductor industry

Status
Not open for further replies.

56860

Senior Member
Registered Member
Just to be blunt despite all the talk in this thread we remain yet to see China come up with a domestic DUV capability. If they had one there’s no reason to keep it so quiet … there’s plenty of demand.
I know, right? Every other country goes 14, 12, 10, 8 progress every month. Why's China so slow? India's doing potato chips, corn chips, almost at yam chips already and China still doesn't even have its first commercial EUV after 2-3 years. It's crazy slow.
 

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
I expect the next step will be the ban of TSMC advanced foundry services for Biren, Baidu, Alibaba and siblings. It is a necessary step to avoid that the banning of NVIDIA and AMD backfires hugely, as many people here have already pointed out.

Banning TSMC and Samsung for serving Chinese customers is one of the main reasons of Chip 4 Alliance existence and the only thing, together with the banning of ASML EUV, that can be effective in forcing, by 2024-2025, a 2-3 generations gap between Chinese and Western companies.

I really hope that SMIC new 7nm capacity will be big enough, because very soon it will be greatly needed.

After US killed the future of AMAT, LAM Research and in general of US semi equipment in Chinese market, this will mark the end of NVIDIA too. But US administartion seems to have already accepted and even embraced the idea of total decoupling in semiconductor and digital market.
The good thing is that US politicians normally don't function that quickly. They will probably give a year or two before they realize Chinese alternatives are available and that they just torpedoed Nvidia/Qualcomm/AMD.

The key for all the Chinese AI chips is that they can be done with 7 nm process, which is something SMIC has already demonstrate with their N+1 process.

Aside from Baidu, the one to watch out for is HiSilicon
Please, Log in or Register to view URLs content!

A lot of their newer chips just couldn't be produced in the recent years. @latenlazy mentioned this to me, but Huawei probably doesn't care about high margins/yield on its 7 nm or 14nm process, so they might just be able to crank out a lot of AI/server chips with low margins for a while. Everyone else can keep using TSMC/Samsung process until they get banned or domestic ramp up. After all, they've already expanded a lot of energy in designing their chip. It's more important for them to have production ramped up first.

SMIC needs to be buying as many ASML/SMEE DUVs as possible. I'd be curious to see the full year import number from ASML. This is time for Chinese government to step in and support construction of new fabs so they can meet local demand. There probably has never been a more optimistic time if you are a chip maker in China.

I tried to find an alternating source for this info and stumbled across this Bloomberg article

Please, Log in or Register to view URLs content!

And among other things it says: "ASML is investigating the possibility of increasing shipments to 90 of its EUVs in 2025, from an original target of 70 units"

Now everybody is buying / planning to buy EUV (Intel, Samsung, TSMC, Micron, etc), but by 2025 /2026 ASML will face a serious overcapacity issue because the fabs now planned will be already built, and very probably there will be a huge overcapacity in semiconductor market and semi equipment makers will feel the pain the most, as always happens: machinery manufacturers are the first to suffer a customer's market downturn because investments in new capacity are the firsts to be scrapped /delayed.

In view of this, it is strategic critical for ASML to keep the gates with China open, or it will pay a very high price after 2025-2026. I guess in ASML they have this thing very clear, and is one of the reasons why although the ultra-aggressive pressure, US have still not succeeded to force ASML banning DUV in China.
I think there will be a huge requirement from China for their DUVs over the next couple of years. After that, ASML will be in trouble in China unless they start exporting EUVs. This is a conversation ASML needs to be having with Dutch government. At the end of the day, US government is already forcing the decoupling of its own firms with China. The next step is force decoupling of Samsung and TSMC from China. So while ASML probably does get most of its revenues from Taiwan and SK right now, that won't be the case if China is decoupled from TSMC/Samsung.
 

tokenanalyst

Brigadier
Registered Member
I tried to find an alternating source for this info and stumbled across this Bloomberg article

Please, Log in or Register to view URLs content!

And among other things it says: "ASML is investigating the possibility of increasing shipments to 90 of its EUVs in 2025, from an original target of 70 units"

Now everybody is buying / planning to buy EUV (Intel, Samsung, TSMC, Micron, etc), but by 2025 /2026 ASML will face a serious overcapacity issue because the fabs now planned will be already built, and very probably there will be a huge overcapacity in semiconductor market and semi equipment makers will feel the pain the most, as always happens: machinery manufacturers are the first to suffer a customer's market downturn because investments in new capacity are the firsts to be scrapped /delayed.

In view of this, it is strategic critical for ASML to keep the gates with China open, or it will pay a very high price after 2025-2026. I guess in ASML they have this thing very clear, and is one of the reasons why although the ultra-aggressive pressure, US have still not succeeded to force ASML banning DUV in China.
I think ASML plans is to have high NA EUV ready for 2025 and pressure the Dutch goverment to be allowed to sell low NA scanners to China before they make their own EUV scanners under the reason that low NA EUV is "mature technology".
 

Zichan

Junior Member
Registered Member
I think ASML plans is to have high NA EUV ready for 2025 and pressure the Dutch goverment to be allowed to sell low NA scanners to China before they make their own EUV scanners under the reason that low NA EUV is "mature technology".
China is not able to manufacture DUV systems even remotely competitive with anything ASML sells. While ASML would undoubtedly like to sell its EUV systems to Chinese companies, they most certainly aren't losing any sleep over worries about a potential Chinese EUV rival in the next 5 years.

There's no economic way to leapfrog DUV and go straight into EUV. It took ASML 10 years to go from a EUV prototype to a machine capable of high volume manufacturing. By the time they made the first prototype they were already a market leader in DUV immersion. There are simply too many critical technologies that need to be mastered on DUV before being able to move on to EUV that make leapfrogging attempts a huge risk.
 
Last edited:

pbd456

Junior Member
Registered Member
If Gigaphoton develops their EUV light source the full responsibility will shift to the Japanese and Dutch goverment.
Or if China develops their own light source the responsability will be shifted to only the Dutch goverment.
ASML signed agreement to have a certain percentage of US tech inside EUV, not to mention what US could do to Netherland.
 
Status
Not open for further replies.
Top