What you are speaking about is plasma strip systems, which are called degumming in China. According to my analysis at The Information Network (), PSK (Korea) is the market leader with a 32% share followed by Mattson Technology at 25% in 2021. Mattson was taken over in 2015 and the Chinese name is Yitang Semiconductor.
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
It continues to amazes me that the politicians are so obsessed with Huawei and its factories while oblivious to how well adjusted the rest of Chinese IC industry is. SMIC is the real champion.
Anyways, saw this morning bout the top 14 largest SMIC clients. I'm not sure if this is true or how old this is the data
#1 is Qualcomm, unexpectedly and is apparently 10 to 12% of SMIC revenues. They even do low end phone SoC (using 14nm process) for Qualcomm here.
#2 is Broadcom, 3 to 7% of revenue. RF ICs with 8 inch wafers
#3 is Gigadevice , 4 to 6% of revenue. NOR Flash, SLC NAND flash, MCU
#4 is Galaxycore, 2 to 4% of revenue. CIS (8 inch, 12 inch)
$5 is Unisoc, RF IC, Wifi/bluetooth, DPU (12 inch wafer). Surprised that SoC is not listed here.
#6 Will Semiconductor, 1 to 4% of revenue. more CIS (8, 12 inch)
#7 Sanechips, 1 to 3% of revenue
Anyways, saw this morning bout the top 14 largest SMIC clients. I'm not sure if this is true or how old this is the data
#1 is Qualcomm, unexpectedly and is apparently 10 to 12% of SMIC revenues. They even do low end phone SoC (using 14nm process) for Qualcomm here.
#2 is Broadcom, 3 to 7% of revenue. RF ICs with 8 inch wafers
#3 is Gigadevice , 4 to 6% of revenue. NOR Flash, SLC NAND flash, MCU
#4 is Galaxycore, 2 to 4% of revenue. CIS (8 inch, 12 inch)
$5 is Unisoc, RF IC, Wifi/bluetooth, DPU (12 inch wafer). Surprised that SoC is not listed here.
#6 Will Semiconductor, 1 to 4% of revenue. more CIS (8, 12 inch)
#7 Sanechips, 1 to 3% of revenue
Lithography is still a big issue, SMEE can't compete directly with ASML even if their scanners are good enough and looks like ASML will be there.
In one hand SMEE cannot be doing this expensive scanners forever if nobody is going to buy them, they risk losing the packaging market to others new types of lithography machines like UV LED lighting and maskless.
Even CETC automatic mask aligners are going UVLED now.
In the other hand you have ASML, the indispensable company that you need to build a semiconductor industry but they are a giant ticking time bomb, will only take a group of fascists anti-business anti-china hawks to take over Europe and ASML is gone. In fact I think their monopoly is pretty dangerous for the whole semiconductor industry as a whole, if something bad happen to Europe the whole semiconductor industry will be derailed.
So a balance is needed, SMEE needs to share at least some market share with ASML, their scanners need to be incorporated in Chinese fabs probably along ASML ones to be debugged and troubleshoot, so they can keep researching how to make them better.
Link to the bloomberg article that they are squirming about
by 2025, China will need 500 DUV... so SMIC and other foundries will buy enough to survive before CETC and SMEE complete it in 2023
maybe, 90nm lithography can make 28nm chips... China wants to leave the business to foreign companies.. so companies like SMEE don't focus too much on high-end processes.. now they're speeding it up... better late than never.
so can we expect SMEE's 28nm DUVi on production line in 2023 ?
ASML CEO Peter Wennink: “Based on our initial assessment, the new restrictions do not amend the rules governing lithography equipment shipped by ASML out of the Netherlands and we expect the direct impact on ASML’s overall 2023 shipment plan to be limited.”
This for me is the official confirmation US failed to coerce Holland government.
ASML CEO would not be so clear and direct at the official third-quarter reporting meeting with investors if he isn't sure.
For the time being Holland 1 - USA 0 ...but the match is still long....
Honestly I didn't expect Holland and ASML to resist. I've rarely been so happy to be wrong!
I do think that asml would do okay without Chinese sales. After all, it's only 15 to 20% of their revenue. Much less than 30% for amat, kla and lam. The big concern for asml is losing their monopoly. That would be a much largerhit on asml and dutch economy. As such, it makes sense for them to continue to improve their arfi scanners like nxt2100i. That would be needed for smic's advanced node production in terms of improving yield.
Asml only announced their expansion in china recently to 1500 staff. I don't think they would have done that without assurances from Dutch government that they can continue to sell. This is likely the biggest push from us government until 2024 imo. Unless something drastic happens, smic should be able to get all the scanners they need for smsc expansion and possibly even a lot more than that.
Long term, I do wonder if asml will be allowed to sell euv to china if they can find non American replacement to those parts. The threat of a Chinese competitor in that space is looming for asml.
Aren't we at least semi confident that smee is capable of early generation arfi scanner by this point? It seems like they are at least winning more bid. The future of smee is also questionable given the rumor that cetc and Huawei are also looking to build their own scanners. They will all be using the same ecosystem that smee is using. There is no reason to believe that cetc can't produce something better.
I think smee has a nice head start and have large backers now, but they do need to ramp up production and improve their products if they do not want to be ousted by other Chinese companies.
Is SMEE really being tailgated by CETC?
LOL, Is always a good sign when this mainstream media garbage articles never tell their readers how they come up with their figures, like the 7nm thing, not even a shred of evidence.
Interesting
You need immersion to go below 45nm.
They are risking it.
- Status