• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Canon stepping up to ASML

For sure, and despite us gamers always wanting the tiniest lithography for the most FPS, there's lots of applications that don't need to be on the latest, greatest nodes. Tons of chips are embedded devices in appliances, cars, controllers, etc and don't need to be the lowest power usage or highest performance, they can and are produced with cheaper nodes and do just fine. Plenty of demand for that kind of stuff.
 
Incredible news.

I'm not a Canon user anymore, but I have always respected the technology aspect of the company. I suppose it was about time they expanded into this sector.
 
Sycraft said:
For sure, and despite us gamers always wanting the tiniest lithography for the most FPS, there's lots of applications that don't need to be on the latest, greatest nodes. Tons of chips are embedded devices in appliances, cars, controllers, etc and don't need to be the lowest power usage or highest performance, they can and are produced with cheaper nodes and do just fine. Plenty of demand for that kind of stuff.
Click to expand...

Is there much of a market for new hardware for older nodes? I know there was a burst of spending for them mid-pandemic during the worst of the supply chain crunch; but the impression I got was that was very much outside the norm and old nodes were produced using equipment purchased when they were new until it wore out and they were discontinued.
 
DanNeely said:
Is there much of a market for new hardware for older nodes?
Click to expand...

If only a single generation old for 10% of the price, I would think so. The details are important, (edit: oops pressed submit too early and then couldn't load to fix it), if it's 10% of the price and 5% of the throughput, that won't work for most people, etc.
 
Last edited:
DanNeely said:
Is there much of a market for new hardware for older nodes? I know there was a burst of spending for them mid-pandemic during the worst of the supply chain crunch; but the impression I got was that was very much outside the norm and old nodes were produced using equipment purchased when they were new until it wore out and they were discontinued.
Click to expand...
Is if it is cost effective. There's plenty of stuff that cost for the chip is the biggest factor. So if you can get a chip for $10/chip on TSMC 3nm using ASML or $2/chip on Global Foundries usings Canon 5nm, you take the cheaper one because it is good enough. This gets more true as the new nodes get more and more expensive. While I'm sure the biggest supply will still be older fabs that are still running, I can see demand for things even if not the latest greatest.

Even more so because the improvements with new "nodes" have gotten less because the bullshit has grown. TSMC's 3nm isn't actually 3nm. The marketing term has gotten decoupled from the actual size for awhile now. Hence most companies aren't jumping on it, because of the cost, even though it is the best, you still see plenty of current products on their 5 and 7nm nodes. It IS smaller and it IS better... but not enough to warrant companies jumping straight on it for high performance parts because the uplift isn't enough to justify the cost.

Also Canon's stuff could become more relevant because it sounds like Intel bought ALL of ASML's newest stuff. They threw their money around and bought up their whole production for a bit, so anyone else has to get in line. If Canon is selling something good enough, maybe that is of interest.

We'll see but I think there will be demand. There just isn't the need to be absolute smallest node all the time, even for some high performance components.
 
Sycraft said:
Is if it is cost effective. There's plenty of stuff that cost for the chip is the biggest factor. So if you can get a chip for $10/chip on TSMC 3nm using ASML or $2/chip on Global Foundries usings Canon 5nm, you take the cheaper one because it is good enough. This gets more true as the new nodes get more and more expensive. While I'm sure the biggest supply will still be older fabs that are still running, I can see demand for things even if not the latest greatest.

Even more so because the improvements with new "nodes" have gotten less because the bullshit has grown. TSMC's 3nm isn't actually 3nm. The marketing term has gotten decoupled from the actual size for awhile now. Hence most companies aren't jumping on it, because of the cost, even though it is the best, you still see plenty of current products on their 5 and 7nm nodes. It IS smaller and it IS better... but not enough to warrant companies jumping straight on it for high performance parts because the uplift isn't enough to justify the cost.

Also Canon's stuff could become more relevant because it sounds like Intel bought ALL of ASML's newest stuff. They threw their money around and bought up their whole production for a bit, so anyone else has to get in line. If Canon is selling something good enough, maybe that is of interest.

We'll see but I think there will be demand. There just isn't the need to be absolute smallest node all the time, even for some high performance components.
Click to expand...
So I have to wonder. Is Nikon going to revive its litho equipment? We don't hear much about Nikon anymore in this space.
 
philb2 said:
So I have to wonder. Is Nikon going to revive its litho equipment? We don't hear much about Nikon anymore in this space.
Click to expand...
Nikon has been using Sony sensors for at least 10 years. I don’t think that they have kept up at all on any modern lithography.
 
philb2 said:
So I have to wonder. Is Nikon going to revive its litho equipment? We don't hear much about Nikon anymore in this space.
Click to expand...
Nikon in in a weird place, they now make much more money selling components to other companies to be used in their lithography machines than they make building their own ones. They sort of mothballed their own division on it in 2017 because of a lot of licensing lawsuits, started back up in 2019, and only a few months ago announced they were getting back in the game and have some pretty impressive sounding things scheduled for the Chinese based fabs.
But ASML is still the leader with some 80% of all chips being printed on one of their machines.
 
Would be interesting if canon returns to leading node lithography equipment. Maybe nikon will come back too.
 
Lakados said:
Nikon in in a weird place, they now make much more money selling components to other companies to be used in their lithography machines than they make building their own ones. They sort of mothballed their own division on it in 2017 because of a lot of licensing lawsuits, started back up in 2019, and only a few months ago announced they were getting back in the game and have some pretty impressive sounding things scheduled for the Chinese based fabs.
Click to expand...
Who knew? They are flying under the radar here.

Lakados said:
But ASML is still the leader with some 80% of all chips being printed on one of their machines.
Click to expand...
No doubt. And yes, they could certainly use some effective competition.
 
philb2 said:
I had thought that Canon isn't doing lithography, but "imprinting" wafers.
Click to expand...
That’s just a different form of lithography. But yeah, it’s cheaper and easier to imprint them, but prone to contamination so more pre and post work required between batches so it doesn’t ramp up to scale like the ASML equipment does.
 
Lakados said:
That’s just a different form of lithography. But yeah, it’s cheaper and easier to imprint them, but prone to contamination so more pre and post work required between batches so it doesn’t ramp up to scale like the ASML equipment does.
Click to expand...
Just guessing here, but maybe ASML decided that the contamination issues were too much of a negative.

Really stupid question: Why hasn't a Chinese company tried to use the Canon approach? (Will Canon be allowed to sell to China?)
 
philb2 said:
Just guessing here, but maybe ASML decided that the contamination issues were too much of a negative.

Really stupid question: Why hasn't a Chinese company tried to use the Canon approach? (Will Canon be allowed to sell to China?)
Click to expand...
They do, it works incredibly well at the 32 and 22nm scale.
Canons contribution here is scaling it from 22nm down to 5nm.
LEDs, Solar Cells, MOSFETs, OFETs, all primarily made using this method.

Contaminants aren’t a big issue but but that’s at 22nm, I have no clue how any of this works at 5nm. I’m still trying to wrap my head around how they would make the imprint.

ASML went with DUV because it shrunk past 22nm where the imprint methods got stuck there.
 
Last edited:
You must log in or register to reply here.
Back
Top