I wonder maybe CETC are more advanced than smee
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
Have Chinese manufacturer submitted samples for certification? There is a standard for semiconductor grade polymer components: SEMI F57. Anything that passes is semiconductor grade.
That I don't know but this news came out a few days ago and I keep forgetting to report it
CCCC partnered up with Phytium & Galaxy Kylin to develop a ship Beidou position monitoring system. So looks like Phytium has done quite well with the transportation industry.
Starting for 2021, CEC & CCCC worked together for this project and involving Phytium. This system went online in June of 2022.
Again, I think Phytium & Hygon have gotten many orders because they have been the only desktop CPU makers that had competitive products from last year. It takes time for different ministries to certify new CPUs. This is where Loongson really suffered in not having a competitive CPU like 3A5000 or full software support for LoongArch until this year. But now that Loongson is in a better place and coming out with 3A6000 soon, Phytium still hasn't come out with a competitive product to that. Hygon's 3rd generation has come out but is limited by its license from getting to Zen 3 or 4 level. Sure, Phytium can add to instruction set (don't know if Hygon can), but it will need to get software developers to accept those additional instructions. There is a lot of limitation to that.
With the growth of high-end test orders, Weitest Technology's (V-Test semiconductor) net profit in 2022 is expected to increase by 77.79% to 92.93%
According to news from Jiwei.com, on January 20, Weitest Technology released the 2022 annual performance forecast. According to preliminary calculations by the financial department, it is estimated that the net profit attributable to the owners of the parent company in 2022 will be 235 million to 255 million yuan, which is the same as that of the above Compared with the same period of the previous year (statutory disclosure data), it will increase by 102.8244 million yuan to 122.8244 million yuan, an increase of 77.79% to 92.93% year-on-year. The net profit attributable to the owners of the parent company after deduction of non-recurring gains and losses is RMB 198.000 million to RMB 212.000 million, which will increase by RMB 70.4028 million to RMB 84.4028 million compared with the same period of the previous year (statutory disclosure data), an increase of 55.18% to 66.15%.
Weitest Technology said that in 2022, the company will continue to expand high-end test production capacity, continuously improve test quality and service quality, and at the same time increase the development of new customers. The growth of high-end test orders has kept the company's revenue and net profit at a rapid pace increase.
According to the data, Weitest Technology is a well-known third-party integrated circuit testing service company in China. Its main business includes wafer testing, finished chip testing and supporting services related to integrated circuit testing. The types of wafers and finished chips tested by the company cover CPU, MCU, FPGA, SoC chip, radio frequency chip, memory chip, sensor chip, power chip and other chip types, and the technology covers 6nm, 7nm, 14nm and other advanced processes and 28nm The above mature processes cover mainstream products such as 12 inches, 8 inches, and 6 inches in terms of wafer size, and include communication, computer, automotive electronics, industrial control, consumer electronics and other fields in downstream applications.
(Bolded by me)
1st bold) I believe Hygon is still with it's first generation cpus - their website advertises the 7000, 5000 and 3000 (all based on Zen 1 cores), are you sure you're not thinking of Zhaoxin which actually has had multiple generations of architecture, and has shown improvement, (albeit slower than I'd have liked). Looking at the article in Aug 2022 discussing it's listing in the STAR market (), they appear to be still using the 7000, 5000 and 3000 designs for their cpus.
2nd bold) Also, I don't believe Hygon can legally make changes to the AMD contribution - afaik the zen 1 core (or core complex, not 100% sure) is a separate block IP that can't be altered. Given that AMD can no longer cooperate with the process it means no more core blocks are available. Any modifications without AMD's blessing would be legally impossible (since it'd be altering a component that has no), and prohibitively expensive/risky (as they wouln't have the 'source' files for the
Legally speaking, I'm not sure that any new CPU designs (with an unchanged Zen 1 block) could be made anyway (afaik the design needs to be approved by AMD in the joint venture). If they can self-approve new cpu designs (without changing the core) they might be able to make a SOC with the Zen 1 block and a non-X86 block, though the only thing they could do to maybe stand out against the non-X86 competition (if they want to retain use of the X86 block) would be to make an SOC that contains the Zen 1 block along with non-X86 processing blocks (either RISC-V core, or more logically a gpu block).
Honestly, if they wanted to stay in the X86 sphere they'd be better off investing in Zhaoxin (though that'd be a couple of years to get something significant enough to be worth replacing the 7000 series). That's why I've warmed up to Longsoon (and a bit sad that Zhaoxin hasn't advanced as much as I would have liked).
I think it's just that twitter user got certain "status" with twitter. There is nothing sensitive about the tweet.
I posted about them here Chinese semiconductor industry and Chinese semiconductor industry. They claim to have completed 3rd generation CPU development and have full IP. I don't know what kind of restrictions they have though.
In light of what we know about their AMD licensing, I'm confused on what exactly they can do with their current license. I just assume that they are restricted from utilizing AMD Zen 2/3/4 technology. But in terms of what they can modify to create 3rd or 4th generation Hygon CPUs, I have no idea.
袁琛浩：什么时候SMEE也可以大规模扩产？现在国内几乎所有设备厂商都实现了扩产
havok：其实也有扩产，前面提到的500系列和800系列的2条产线就是扩产的。建一条前道光刻产线的投资可不小，如果投资10亿建一条的话年产可能不到10台，因为还需要配套买若干台先进涂胶显影机、光刻测试用的硅片、各种测试用的光刻胶、各种量测设备套刻检测设备，测试用的准分子激光器、浸没液纯水处理设备等等。
快易典：其它设备咋样啊，中微的高深宽比刻蚀机今年能量产嘛
havok：我只知道中微的临港刻蚀机基地今年下半年量产，在嘉定ICRD那里有3.5亿订单。不过要赶上Lam的水平没那么快
袁琛浩：When will SMEE also be able to expand production on a large scale？Now almost all equipment manufacturers in China have expanded production line
havok：There are production line expansions. The two production lines of the 500 series and 800 series mentioned above are expansions. The investment to build a front-end lithography production line is not small. If you invest 1 billion to build one, the annual output may be less than 10 units, because you need to buy several advanced glue-coating and developing machines, silicon wafers for lithography testing, and various Photoresists for various tests, overlay testing equipment for various measuring equipment, excimer lasers for testing, pure water treatment equipment for immersion liquid, etc.
快易典：How about other equipment? Is AMEC’s high aspect ratio etching machine capable of producing energy this year?
havok：I know that AMEC's Lingang etching machine base will be mass-produced in the second half of this year, and ICRD in Jiading has 350 million orders. But it won't be that fast to catch up to Lam's level
havok：其实也有扩产，前面提到的500系列和800系列的2条产线就是扩产的。建一条前道光刻产线的投资可不小，如果投资10亿建一条的话年产可能不到10台，因为还需要配套买若干台先进涂胶显影机、光刻测试用的硅片、各种测试用的光刻胶、各种量测设备套刻检测设备，测试用的准分子激光器、浸没液纯水处理设备等等。
快易典：其它设备咋样啊，中微的高深宽比刻蚀机今年能量产嘛
havok：我只知道中微的临港刻蚀机基地今年下半年量产，在嘉定ICRD那里有3.5亿订单。不过要赶上Lam的水平没那么快
袁琛浩：When will SMEE also be able to expand production on a large scale？Now almost all equipment manufacturers in China have expanded production line
havok：There are production line expansions. The two production lines of the 500 series and 800 series mentioned above are expansions. The investment to build a front-end lithography production line is not small. If you invest 1 billion to build one, the annual output may be less than 10 units, because you need to buy several advanced glue-coating and developing machines, silicon wafers for lithography testing, and various Photoresists for various tests, overlay testing equipment for various measuring equipment, excimer lasers for testing, pure water treatment equipment for immersion liquid, etc.
快易典：How about other equipment? Is AMEC’s high aspect ratio etching machine capable of producing energy this year?
havok：I know that AMEC's Lingang etching machine base will be mass-produced in the second half of this year, and ICRD in Jiading has 350 million orders. But it won't be that fast to catch up to Lam's level
晓得：我想问使用国产光刻机的28nm制程什么时候量产
havok：算上今年要稳定量产还要2，3年左右，这不单单是光刻机的事，整条线大家共同努力的结果。当然有华为参与可能速度会快些。
bbb aaa：这几天传闻说长存裁员，可能生产难以为继，这靠谱吗
havok：让子弹飞一会儿
晓得：I would like to know when the 28nm process using domestic lithography machines will be mass-produced
Havok：Counting this year's stable mass production, it will take about 2 or 3 years. This is not just a matter of the lithography machine, but the result of the joint efforts of everyone in the entire line. Of course, with Huawei's participation, the speed may be faster.
bbb aaa：There are rumors in the past few days that long-term layoffs may make production unsustainable. Is this reliable?
havok：Let the rumor fly for a while
havok：算上今年要稳定量产还要2，3年左右，这不单单是光刻机的事，整条线大家共同努力的结果。当然有华为参与可能速度会快些。
bbb aaa：这几天传闻说长存裁员，可能生产难以为继，这靠谱吗
havok：让子弹飞一会儿
晓得：I would like to know when the 28nm process using domestic lithography machines will be mass-produced
Havok：Counting this year's stable mass production, it will take about 2 or 3 years. This is not just a matter of the lithography machine, but the result of the joint efforts of everyone in the entire line. Of course, with Huawei's participation, the speed may be faster.
bbb aaa：There are rumors in the past few days that long-term layoffs may make production unsustainable. Is this reliable?
havok：Let the rumor fly for a while
What kind of volume production is CETC looking at though? From what I understand, CETC operates the full stack but as an IDM and makes its own chips, equipment but is it comparable to foundry volumes? I wonder if that expertise is transferrable/scalable or if it is a specialty process unique to CETC that allows them to satisfy their needs but may not be HVM ready currently.
- Status