So to summarise recent events: SMIC is advancing really fast but it's still dependent on foreign DUV machines. SMEE is not advancing as fast as we wish it to be. China is independent from 40 nm and up on its own DUV machines. SMEE's 28 nm DUV machine has not gone as well as planned and it's currently not yet available.
So while we talk a lot about SMIC we should really talk about SMEE.
It is going exceeding well.
This is the problem that needs to be solved, and it is a multi-level problem.
SMEE or any Chinese IC builder of machines, they have to do two things.
1. is build the machine.
2. is to know how that machine works, the one they had not built yet
So how to build a machine that you do not even know how it works?
Obviously, we cannot do that.
There is where ASML old machines come in.
There is the blueprint, just copy that.
So, learn how to use a DUV machine from ASML to use it like a novice, making a 90nm chip, then like a superuser making a 7nm chip. That they already done. The machine is right in front of you, now time to build your own.
But these are complex machines, that must be integrated into current systems inside operating factories.
Hence, there will not be many orders for this SMEE machine at the beginning.
SMEE apparently only shipped 3 of their most advanced machines for testing, to three different clients. More clients, the faster they can work out the bugs. With few clients, it is a slower process, but necessary because of the tech war.
So is apparently where we are in China.
They are testing a domestic IC machine free of any foreign parts.
When they make it work at 100%, which is a certainty, then they will scale up production, and start replacing whatever needs to be replaced.
In a way, this is like saying it is the beginning of the end. It is actually happening once that SMEE machine gets shipped from their factor to a fab for testing. This is version 1 of the manual, and at version 2 or 3 of the manual (after testing), that is it.
Actually, there is a more important part I feel.
This is kind of like the first attempt at such a complex machine for a Chinese company in IC.
Once they make it work, to borrow an expression, then the sky is the limit. Once you break through like this, then they will keep on going.
That know how built up, will just push things along further and faster.
ASML their DUV machines served as a template to go forward.
Once you build your own, you have your own template to go forward, and that probably always is more valuable. Just my opinion.
That is why I would love to see that synchrotron blaster for EUV in China. Blast the buttz!