Chinese semiconductor industry

Status
Not open for further replies.

paiemon

Junior Member
Registered Member
Alright, a little more on Huawei and the domestic chip. The insider that posted it (well at least he is known to have good info in the past) has already deleted that post. I'm sure that attracted more attention than he wanted.

In that same post's comment, he mentioned that 7nm chip will be available for Huawei by 2025. 7nm is basically SMIC's N+2 process and close to the earlier Samsung 5nm in performance IIRC. As such, a stacked SMIC 7nm chipset would be quite advanced. It would also indicate to me that SN2 production has been raised by then. I cannot imagine a scenario where smartphone CPU would be among the first items to be allocated capacity for SMIC's SN2 plant. That tells me we should be seeing N+2 production starting this year for chips to be released on the market by 2024. The initially chips imo will be server chips and maybe some desktop CPUs. Once they've mastered the production techniques, they'd be able to move to more advanced GPUs and smartphone chips by sometimes in 2024 or early 2025.

Now today, I saw a post where someone asked him about 12nm 5G chip again. In a rather shy response, he pointed to Huawei Enjoy 50z. From what I can see, it makes no mention of CPU and doesn't support 5G
Please, Log in or Register to view URLs content!
.
Please, Log in or Register to view URLs content!
Unlike Huawei Enjoy 50 which is stated to use Kirin 700 series. There is nothing listed for 50z from what I can see.
That would suggest to me that a different version of this will come out in 2023 that will be stacked 12nm and support 5G. Again, I don't see anyone other than SMIC for this.
Interesting observation. Would this mean that SN2 is dedicated to N+2 and the existing SN1 is dedicated to 14/12/N+1 then?
 

tokenanalyst

Brigadier
Registered Member

Beijing Shuoke Zhongkexin (CETC semiconductors) signed a total of 11 new machines (ion implanters) at the beginning of the year, and the total contract value exceeded 200 million yuan​


Jiweiwang news, on January 9, Beijing Shuoke Zhongkexin Electronic Equipment Co., Ltd. (hereinafter referred to as "Beijing Shuoke Zhongkexin") announced that a total of 11 new machines were signed at the beginning of the year, and the total contract amount exceeded 200 million yuan.

It is reported that Beijing Shuoke Zhongkexin was established in 2019. The company originated from the 48th Research Institute of CETC. It is an earlier high-end equipment supplier in China that focuses on the ion implanter business in the integrated circuit field. In the 1960s, it began to study ion implanters, and successively undertook the research tasks of major projects such as 90-65nm medium beam current and 45-22nm large beam current. The four products of ion implanter and customized ion implanter have built a relatively systematic pedigree of integrated circuit ion implanters.

In August 2022, Beijing Shuoke Zhongkexin's fourth medium-beam ion implanter was successfully Move-in in an integrated circuit production line; in September of the same year, Beijing Shuoke Zhongkexin medium-beam ion implanter was successfully delivered.

Please, Log in or Register to view URLs content!
 

tphuang

Lieutenant General
Staff member
Super Moderator
VIP Professional
Registered Member
Could you explain why you infer this to be the immersion stage instead of the dry stage?
I don't have the answer to that one. Simply relaying the original poster. My guess is their product can be used in both Arf dry and immersion machines as @tokenanalyst posted below.

8 deliveries from 2020 to 2022 for 8 prototypes of both dry & immersion at different stages of development cycle seems reasonable figure.
The 1nm precision match with their statements about the dual maglev stage in their documents of dec. 2021.
Also kind of interesting that in 2017 when the immersion machine was planed for 65nm they stated that their dual maglev stage was going to be used in KrF and ArF dry, ArF Immersion and I guess it could be use in future EUV machines.

View attachment 104633
Right, no need for people to panic about U-Precision. It seems like they set out to do this from a few years back and they have delivered quite a few of their product already and they will be ready to deliver more once the testing & validation is finished. Just need to ge the funding to ramp up.

Interesting observation. Would this mean that SN2 is dedicated to N+2 and the existing SN1 is dedicated to 14/12/N+1 then?
Well, I think it's safe to just assume that most if not all of SN1 is dedicated to 14/12/N+1 and probably heavily tilted to 12 at this point. Others would know better than I do here. If 12 is just an improvement over 14, how much work is required to convert the 14 nm to 12nm process or is that not worth the time? It seems like a lot of the chips we are discussing recently are using 12nm process. I think N+1 right now is still new and they are still working up the yield and complexity on them. If say 10k out of 35k wpm are N+1, then its likely they can still increase throughput on that capacity as they improve their process.

I have no idea about SN2. We will see if they can get the tools they need. My impression has been that finding tools even to fully buildout SN1 was not easy. And that lithography machines are not the missing piece. There are other tools that are hard for them to acquire. I'd imagine it would be far easier for them to put in additional 12nm and N+1 capacity than N+2. But reading along the grapevine, it does seem like they are still seeing a path to produce N+2 chips over the next 2 years.
 

BoeingEngineer

Junior Member
Registered Member
i have doubt that, South korea produce anything related critical chip tools. they itself highly dependent on foreign made tools.

Please, Log in or Register to view URLs content!

China indigenous tools percentage is highest among these countries. that too as of 2021.

Remember that was back in 2021 !!

China is making progress in the matter of weeks !!
 

Orthan

Senior Member
Bloomberg article about the US government sanctions against china´s semiconductor industry.

Please, Log in or Register to view URLs content!


Its largely behind a paywall. For those that can acess the whole article, does it present new information not yet posted on this thread?
 

daifo

Captain
Registered Member
Bloomberg article about the US government sanctions against china´s semiconductor industry.

Please, Log in or Register to view URLs content!


Its largely behind a paywall. For those that can acess the whole article, does it present new information not yet posted on this thread?
For academic purposes... article is nothing new and has similar commentary posted many times over

Please, Log in or Register to view URLs content!
 

tokenanalyst

Brigadier
Registered Member
For academic purposes... article is nothing new and has similar commentary posted many times over

Please, Log in or Register to view URLs content!
Yeah, as I suspected, Xi Jinping blah blah blah Biden blah blah sanctions blah blah blah YMTC blah blah corruption blah blah blah allies blah blah blah. in summary the same robotic garbage in repeat. What a shock.
 

gelgoog

Lieutenant General
Registered Member
Always the same retarded droning in that Bloomberg article I agree. I will just add that Tsinghua's Unigroup's former CEO was not just responsible for disasters like would seem if you read the article. After all they funded and nurtured YMTC into being and that was pretty successful. A bit too successful one would say. I do think Spreadtrum kind of lagged behind after he acquired the company and put it under UNISOC however. MediaTek still has way better offerings. And pretty much everything UNISOC makes which is in demand is made at TSMC which is risky to say the least. UNISOC also has zero RISC-V offerings. I think it is reductive to only design chips for smartphone and tablets, but whatever, more work for Rockchip I guess. Not that Rockchip is much better with them making their chips at GlobalFoundries.

His major downfall was the shady land deals he made for those NAND and DRAM fabs Unigroup were supposed to build, other than YMTC, which never went anywhere. I also think he spent on way too diverse investments and to a large degree his business model makes little sense. I was never much of a fan of conglomerates. Typically lead to mediocrity.
 
Last edited:
Status
Not open for further replies.
Top