Just a heads up. This is an English speaking forum, so you are gonna have to type in English.
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
Hi, the rules of the forum are English language.
The problem is not the microarchitecture, but the manufacturing itself. In fact there are a number of Chinese companies such as Xiaomi who are working on RISC-V products. I believe Huawei has the full architectural license from ARM, so even with the sanctions in place, they can still produce the designs. However, they still need a factory to produce it, and TSMC is the only one who can produce the cutting edge designs.
A lot of Chinese tech companies are actually committing themselves to open-source software and hardware for precisely the reason you mentioned, independence from USA. My own observation is that open-source platforms are also more popular in Europe, however all the big tech companies are American, so it is an uphill battle.
China Still Far From Semiconductor Self-Sufficiency, Report Says
In my opinion the goals of Made in China 2025 for the semiconductor sector (70% made in china) seems impossible, but there will be progress in the years to come (YMTC, CXMT, Naura, AMEC ect)
In my opinion the goals of Made in China 2025 for the semiconductor sector (70% made in china) seems impossible, but there will be progress in the years to come (YMTC, CXMT, Naura, AMEC ect)
Sorry. I am used to writing in French. I repeat my question.
Hi everyone. This is my first intervention at this forum. I have a question. Why is huawei not interested in risc-v open architecture? The arm architecture is not outside the power of the United States and is not reliable and secure in the long term. I know the company is a member of the risc-v community but does it have any plans to design processors?
While waiting for smee 28nm duv certification, what I like to see is forming the "China EUV alliance" to spur the development effort.
Right now it's exclusively fall in the hands of CAS and other research institution.
I think more is needed.
"China EUV alliance" would consists of many industrial players like huawei, zte, smic, Baidu, Alibaba, tencent, gree, byd, cetc, smee, Xiaomi, Lenovo...etc
And all relevant research academic and the Chinese government.
A pool of $5Billion funding is established. Government will fund a portion and start to collect from industrial players
This would expedite CAS EUV development. Government supervisory group will oversee the development effort. Research institutes would need to greatly expand their man powers and resources
More schools and universities are welcomed to join and smaller tasks of this project can delegated to them
Bi -Weekly progress needed to report and emailed to all members.
What I don't want to see it's an underfunded CAS and other research insitututes buried themselves in close developing environment and struggled with slow progress
DUV already near the final industry certification stage and it will run its course and time is needed. Anxiously waiting won't help in the industry. But EUV development needs a jolt.
Right now it's exclusively fall in the hands of CAS and other research institution.
I think more is needed.
"China EUV alliance" would consists of many industrial players like huawei, zte, smic, Baidu, Alibaba, tencent, gree, byd, cetc, smee, Xiaomi, Lenovo...etc
And all relevant research academic and the Chinese government.
A pool of $5Billion funding is established. Government will fund a portion and start to collect from industrial players
This would expedite CAS EUV development. Government supervisory group will oversee the development effort. Research institutes would need to greatly expand their man powers and resources
More schools and universities are welcomed to join and smaller tasks of this project can delegated to them
Bi -Weekly progress needed to report and emailed to all members.
What I don't want to see it's an underfunded CAS and other research insitututes buried themselves in close developing environment and struggled with slow progress
DUV already near the final industry certification stage and it will run its course and time is needed. Anxiously waiting won't help in the industry. But EUV development needs a jolt.
Last edited:
Well it is impressive, I think, that China went from a situation where they basically did not manufacture any memory products worth mentioning to having YMTC and CXMT manufacture Flash and DRAM in particular. People sometimes think these things are easy but Taiwan had at least 3 different DRAM manufacturers even not that long ago and they went belly up. I still remember purchasing memory from Nanya in the early 2000s. They had a competitive product in terms of capacity and price. Still it had its issues. I had to swap the memory sticks twice until I got RAM which worked long term. However they had a fairly liberal returns policy and an extensive warranty so I was fine with it. But that can't be healthy for a company financially and it shows. Infineon/Qimonda (German company) is another example. They used to have a decent amount of market share at one point and now they're dead. They had trouble shrinking down their memory cells and increasing the density of the memory to produce higher capacity chips. To continue required massive investments and time they didn't have. The Japanese had to merge all their memory companies into one to survive. It takes a lot of continuous investment and the price wars, when they happen, can be quite difficult to stomach. They can have vast profits one year to be selling at a loss the year afterwards. Current DRAM technology is also expected to hit a brick wall over the next decade as it is hard to shrink the size of memory cells further without a redesign or change in basic technology.
It will take huge investments in the tens or hundreds of billions over the next decade just to compete in that market.
It was never going to be possible to hit 70% Made in China semiconductors by 2025 without massive foreign investment into manufacturing in China. The deal China made with Qualcomm to fab their SoCs in China was an example of that. But with the recent sanctions on SMIC I bet that went down the drain. A lot of the projects were also bound to fail as many of these companies have limited market experience. Still I think the initiative has done decently enough as, like I said, today China does manufacture many products it previously did not even if the quantities are still limited. In a way I think the sanctions might help since it means when China does ramp up capacity it will use Chinese tools which can't be sanctioned by the US. I think the US has done the sanctions at least a decade too late and it's basically impossible to kill the Chinese semiconductor industry at this point. Had they waited another 2-3 years people would probably not even have noticed the sanctions much at all.
Last edited:
D
Deleted member 15949
Guest
They didn't. The 100% tariff was non-MFN. The Netherlands is MFN
This is the kind of shit that starts thing like pearl harbor.
- Status