Well they promise the King will return, so Long live the King!!!!
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
broadsword
Brigadier
Well they promise the King will return, so Long live the King!!!!
Now I recall two years ago, Meng Wanzhou said 2023 would be a turning point.
Can't wait for next year, I want to know the performance level of new Kirin 9000 series and IF it use SMIC N+3 (5nm) tech.
SMIC N+3 is 6nm node. Its critical feature sizes are much larger than 5nm.
I just published this newsletter on the SMIC/Huawei issue. Let me know your thoughts:
China Has Won The Chip War On Ineffective Sanctions, But ASML Could Be Impacted Says Dr Robert Caste
Huawei's new Kirin chip is manufactured by Chinese company SMIC, marking China's success in the chip war despite US sanctions. It was built using ASML's DUV immersion lithography equipment and the U.S. will further sanction the technology
By:
ALLENTOWN, Pa. - Sept. 6, 2023 - -- News out today that the new Kirin chip is manufactured by Chinese semiconductor company Semiconductor Manufacturing International Corporation (SMIC). It is Huawei's first higher-end Kirin processor since 2020 after the U.S. government restricted American businesses from selling their products or services to Huawei.
The Mate 60 Pro, introduced just last week, coincided with a high-profile visit by US Commerce Secretary Gina Raimondo to discuss bilateral business relations. This timing led state media to celebrate the perceived failure of US trade restrictions.
But there are three issues that point me to acknowledge that China has won the chip war:
The Mate 60 is powered by a new Kirin 9000s chip was made on SMIC's 7nm node technology
The Mate 60 was made with a component localization rate of 90%
The Kirin 9000 was made using ASML's DUV immersion lithography systems but domestic equipment is increasing share of semiconductor equipment spend in China.
These two topics are discussed in detail in this article, which are complimentary to my thesis that U.S. sanctions have not been working in China
The ability of China's SMIC to make advanced ICs has been stymie by U.S. sanctions. But the ability for SMIC to reach 7nm is nothing new. I pointed that out 15 months ago in a May 18, 2022 Newsletter entitled "Applied Materials: SMIC Move To 7nm Node Capability Another Headwind," which can be found on my website.
Importantly, in my newsletter I noted:
"My analysis shows that SMIC is in production on its N+2 node, equivalent to the 7nm node and comparable to TSMC. And it was accomplished without the use of EUV lithography, just as TSMC did with its first generation 7nm node (N7) introduced in 2018.
Instead of SMIC manufacturing 28nm logic chips at its new fabs in Shanghai and Shenzen China, SMIC has moved first to its N+1 node (8nm equivalent) and then N+2 node. Moving to a smaller node requires a different semiconductor processing equipment product mix, benefiting equipment suppliers."
I anticipate that the U.S. will impose these long-anticipated sanctions on ASML's DUV immersion lithography systems. Importantly, SMIC was able to reach 7nm despite U.S. sanctions and despite EUV sanctioned lithography equipment.
Web -
The number of advanced lithography machines currently owned in mainland China:
1. 1980i: 6 SMIC South (2018), 6 SMIC Beijing
2. 2000i: 5 SMIC South (2023/7)
3. 2050i: 4 SMIC South (2 in 2021, 2 in 2023/6: 2)
4. Nikon635: 5 sets of ICRD Jiading in the last episode (2022)
1. 1980i: 6 SMIC South (2018), 6 SMIC Beijing
2. 2000i: 5 SMIC South (2023/7)
3. 2050i: 4 SMIC South (2 in 2021, 2 in 2023/6: 2)
4. Nikon635: 5 sets of ICRD Jiading in the last episode (2022)
This topic is actually one I have some specific knowledge about since the instruments I'm currently working with also use droplet generators (but not for molten tin). The key question to center any analysis of performance parameters around here is *what kind of droplets and stream do you want*. It may simply be the case that SIOMP *wants* larger droplets at higher frequency, perhaps based on however they've approached excitation interaction between the laser pulse and droplet. The droplet size is determined by the stream diameter which itself is determined by the inner diameter of the channel which your stream is flowing through, which is very easy to control, so imo it's not really a meaningful performance parameter to draw comparisons around. Similarly spacing constraint as ZeEa5kPul noted is frequency dependent, *but* also stream velocity dependent. Same frequency with higher stream velocity means greater spacing. This is also a very easy parameter to control, and the droplet spacing choice probably has more to do with what stream velocity they want as determined by physical configuration and/or dimensions of their excitation target area and perhaps the debris collection system. Probably the only two metrics that can provide meaningful performance comparison are the position instability and operation time, but depending on other differences in how they might have optimized the excitation interactions the actual differences in those numbers may also not matter. Without understanding *very* specific details about the whole system of interactions for each instrument it's very hard to draw meaningful performance judgments just by looking at these kinds of numbers (this is why you need systems engineers folks).
Huawei has two 1980di.
@hvpc this is the first time that I read about SMIC N+3. I just searched the net but absolutely no info on this.
Can you share a bit more about SMIC N+3? Is there any public info on this? Or is something you know from your insider info?
- Status