Last year I heard that the I-line dual wafer stage has reached 150-200 WPH, how far is that from HVM?
Chinese semiconductor industry
- Thread starter Hendrik_2000
- Start date
- Status
in most things, yes. But if they have put SMEE DUVi machine in high volume production, I think you will see some Reuters or SCMP article out about it for certain industry sources. Like what you saw with YMTC news recently.
havok already said this. SMEE immersion machine is still at testing phase
havok：浸没式量产还早呢，上微去年还亏损呢 It’s still early for immersion mass production
@olalavn today posted some information regarding SMEE's DUVi serial production
he says 2024-25.
he also said, dual layer is very hard so machine is continuously in testing phase as of now.
Last edited:
In his view if it isn't exactly a drop in replacement for ASML tools it might as well be stone age useless garbage no matter what else.
It's not my view. It's every fab's view.
I'm not the one deciding whether or not to put the tool into production.
Fab care about Good Wafer Out. A machine that "runs" (passes all standard tool acceptance spec) but can't consistently pump out good wafers is not useful to any fabs.
Havok said SMEE has a funding issue，as they are considering to go public to raise funds.
I think Chinese gov should grant SMEE enough fund to produce more DUVi prototypes，in order to accelerate the testing process，it would not be cheap，but I think it's worth.
I guess the only way to compel Chinese gov to do that，is ASML stop selling DUVi to China
I think Chinese gov should grant SMEE enough fund to produce more DUVi prototypes，in order to accelerate the testing process，it would not be cheap，but I think it's worth.
I guess the only way to compel Chinese gov to do that，is ASML stop selling DUVi to China
I'm not really sure why I was replied to there.
SMIC & ICRD are validating Arfi right now based on Havok's comments. SMEE seems to be incapable of handling more than 1 project. Let's if they can get SSA800 up to speed first.
If SMEE scanning speed & yield is not up to par (which most certainly is the case at this point), then Chinese gov't can compensate SMIC for that. There has to be a capital cost that makes up for the headaches of dealing with a less productive/more troublesome machine. This is not exactly new for Chinese companies. Many other industries have had to deal with using inferior Chinese tools until they became competitive.
I would imagine SMIC, ICRD & YMTC all are testing/validating SMEE Arfi scanners at this time. It seems to me Chinese gov't should just compensate the fabs themselves for using the tools by paying them out of pocket
Yeah I get it. But you know what is more important to a Fab than good wafer out? Not dying. IMO having a inferior tool is better than having no tools at all, at least is something that can be improve upon. SMIC and others don't know far this is going to escalate or if ASML will be allowed to export any tools at all in the future and the only way SMEE can improve their tools is by the grueling trial and error of a real fab experience. Chinese fabs need to start using SMEE tools even if the Chinese government have to subsidize the lost of yield. Ironically is the only way they have to guarantee that ASML will be allowed to export their tools in the future.
i m the least technical person here on semiconductor. but wanting to give my opinion based on other high end industries.
this is how your journey begins. initially Chinese machine tools were inferior and accuracy was not upto the mark. errors and cutting process was so bad that military industrial complex outright rejected indigenous tools but KEDE and other firms continue to work on tools with feedback from local customers. and look at now where KEDE stands. not only they solved all technical bottlenecks but replaced most of the core components with local ones. now China accounts more than 50 percent of machine tools patent grants globally according to WIPO. now they are competing with market leaders.
SMEE somehow build the foundation of immersion. as per Havok all technical bottleneck have solved too. so initially if SSA800i has poor performance with low yield. its normal i think. just wait for few years. they will produce fully competent machine in high numbers.
- Status