hckrnws
ASML unveils EUV light source advance that could yield 50% more chips by 2030
by pieterr
This video is a really cool dive into EUV for the uninitiated (me) https://youtu.be/MiUHjLxm3V0?si=kEPSicC2WXYhcQ6L
Or this video, which came out before Veritasium's
Or this Asianometry video which came out even sooner.
Glad to see Branch Education represented here.
Or this presentation which came out way long ago.
This is worth the (re)watch every time it comes up.
"I didn't want my name associated with this on the internet"
I didn’t, still don’t, but that’s a lost cause.
I’ll note that this video is way out of date…both in content and my skills as a speaker :P
Thanks for the informative presentation!
thanks for the HN community - the video is how I ended up here and its one of the few social media-esque sites I bother visiting. Taught me a pile of things about coding and CS that weren't in my mechanical engineering degree.
> Thanks for mentioning ASML sponsoring this. I was about to buy an EUV machine from another vendor
lol
I thought this video was a lot better than the Veritasium video. The Veritasium video was awkward. I think they tried to follow the formula from the (excellent) blue led video that performed so well, but it just didn't work.
Disagree, I thought the Veritasium video was fantastic. You understand how the machine works in depth, the history of its development and challenges it encountered, and hear from people actively working on it. It’s a science lesson and history lesson. Like usual, they keep the video engaging and focused on the story, while still keeping a lot of depth with the science. It’s a great format
The whole “exploding tiny drops of metal” in the middle of this is just Loony Toons. This machine is literally insane and two of the companies I am long-long on would be completely fucked without it.
You forgot WITH LASERS, and IN A VACUUM
IIRC from the Veritasium video[0] there is actually some hydrogen gas flowing at quite a high speed though the laser chamber to carry away the tin debris so that it does not accumulate on the mirrors.
They account to every single tiny atom somehow too, but I think I fell asleep last time I watched the video.
The old SemiAccurate article https://semiaccurate.com/2013/02/13/euv-moves-forward-two-st... was very funny.
Seeing this news story made me briefly fear that they’d found a way to replace this glorious mechanism. Thankfully not. In fact, they’re going to shoot more droplets, more often!
So much more fun than LEDs.
Yes it was crazy when I first heard about it "wait what? they shoot it in mid-air?" and that was before I found out they did that like 30k times a second.
But now 100k times a second apparently. Humans are amazing.
You have a machine that’s basically a clean room inside and one of the parts is essentially electrosputtering tin but then throwing all the tin away and using the EM pulse from the sputter to do work.
Oh and can you build it so it can run hundreds or thousands of hours before being cleaned? Thanks byyyyyyyyeeeeee!
The inside of those machines are far, far cleaner than the inside of any clean room ever entered by a human. They have to be molecularly clean.
Which isn't easy considering they explode tin droplets in the machine. I think that's the point the other commenter wanted to make.
Think about the purity requirements that places on the tin.
> We are going to spray expensive stuff in an extremely fine and precise line. Then we're going to shoot a laser at each droplet.
< Why?!
> To make a better laser.
< Yes, of course you are.
> 100,000 times per second.
< [AFK, buying shares.]
Don't forget that they are hitting each droplet 3 times.
I have shares in one of their biggest customers, and one of their customer’s biggest customers.
We are quickly leaving the realm of dependent variables still looking anything like diversification.
It seems like you want someone to ask you what the two companies are. So - what are the two companies?
Nvidia and an AI company
> We are quickly leaving the realm of dependent variables still looking anything like diversification.
What does that mean?
That is why each machine costs a few hundred million eurodollars.
The thing I didn't understand after watching that video was why you need such an exotic solution to produce EUV light. We can make lights no problem in the visible spectrum, we can make xray machines easily enough that every doctors office can afford one, what is it specifically about those wavelengths that are so tricky.
The efficiency of X-ray tubes is proportional to voltage, and is about 1% at 100kV voltage. This is the ballpark for the garden variety Xray machines. But the wavelength of interest for lithography corresponds to the voltage of only about 100V, so the efficiency would be 10 parts per million.
The source in the ASML machine produces something like 300-500W of light. With an Xray tube this would then require an electron beam with 50 MW of power. When focused into a microscopic dot on the target this would not work for any duration of time. Even if it did, the cooling and getting rid of unwanted wavelengths would have been very difficult.
A light bulb does not work because it is not hot enough. I suppose some kind of RF driven plasma could be hot enough, but considering that the source needs to be microscopic in size for focusing reasons, it is not clear how one could focus the RF energy on it without also ruining the hardware.
So, they use a microscopic plasma discharge which is heated by the focused laser. It "only" requires a few hundred kilowatts of electricity to power and cool the source itself.
The issue isn't in generating short wavelength light, it's in focusing it accurately enough to print a pattern with trillions of nanoscale features with few defects. We can't really use lenses since every material we could use is opaque to high energy photons so we need to use mirrors, which still absorb a lot of the light energy hitting them. Now this only explains why we need all the crazy stuff that asml puts in it's EUV machines to use near x-ray light, but not why they don't use x-ray or higher energy photons. I believe the answer to this is just that the mirrors they can use for EUV are unacceptably bad for anything higher, but I'm not sure
Photoresist too. XRays are really good at passing through matter, which is a bit of a problem when the whole goal is for them to be absorbed by a 100 nanometer thick film. They tend to ionize stuff, which is actually a mechanism for resist development, but XRay energies are high enough that the reactions become less predictable. They can knock electrons into neighboring resist regions or even knock them out of the material altogether.
There is such a thing as X-ray lithography, but it comes with significant challenges that make it not really worth it compared to EUV.
I'd like to hear more about these challenges
Stochastic effects become a bigger and bigger problem. At some point (EUV) a single photon has enough energy to ionize atoms, causing a cascade that causes effects to bloom outside of the illumination spot.
As I understand it, primarly because due to the high energy level of x-rays, light x-ray interacts very differently with materials[1]. Primarily they get absorbed, so very difficult to make mirrors or lenses, which are crucial for litography to redirect and focus the light on a specific miniscule point on the wafer.
The primary method is to rely grazing angle reflection, but that per definition only allows you a tiny deflection at a time, nothing like a parabolic mirror or whatnot.
All of these problems or equivalent still exist in EUV. Litho industry had to kind of rethink the source and scanner because it went from all lenses to all mirrors in EUV. This is also why low NA and high NA EUV scanners were different phases.
As I hear it, the decision had large economic component related to Masks and even OPC.
100%. EUV barely works. XRay litho takes all the issues with EUV and cranks them up to 11. It will take comparable effort to EUV, if not more, to get XRay litho up and running, and I'm not aware of anyone approaching this to anywhere near the level of investment that ASML (and others) have pumped into developing EUV tech. We may get there eventually as a species, but we're a ways off.
If you think it barely works now, you should've seen it when we first started. Availability of a machine was "fuck you"% and the whole system was held together by duct tape, bubblegum and hope. Compared to that the current system is entirely controllable.
Oh, for sure, via herculean effort and investment we have created ourselves a functioning and economical process!
We do actually have functioning processes for XRay litho today, but we'll need that same level (or more) of investment and effort to make it economical.
There are no normal x-ray mirrors. The only way to focus them is to use special grazing mirrors where the x-rays hit them almost parallel to the surface.
https://science.gsfc.nasa.gov/662/instruments/mirrorlab/xopt...
Comment was deleted :(
It really is the specific wavelength. Higher or lower is easier. But euv has tricky properties which make it feasible for Lithography (although just barely it you have a look at the optics) but hard to produce with high intensities.
Specifically, what makes x-rays easy to generate are these: https://en.wikipedia.org/wiki/Characteristic_X-ray In essence, smashing electrons into atoms allows you to ionize the inner shell of an atom and when an electron drops down from an outer shell, the excess energy is shed as high-energy photons. This constrains the energy range of X-ray tubes ("smash electron into metal") to wavelengths well below 13.5nm.
(These emission lines are also what is being used in x-ray spectroscopy to identify elements)
You can also generate broad spectrum bremsstrahlung radiation easily, this is widely used for medical X-rays.
Any source to this? I am hearing this for the first time.
ITs easy to make X-rays, you just hit a metal target with electrons: https://en.wikipedia.org/wiki/X-ray_tube
You can hit metal the same way for EUV.
I assume this doesn't work well otherwise everyone would be doing it.
No you can't, or rather you only get a tiny amount in the correct wavelengths
https://www.youtube.com/watch?v=5Ge2RcvDlgw
Asianometry has lots of videos on ASML, this one is specifically about the light sources.
Here's your link without the surveillance
With slightly less surveillance
Touché. Here's the link without surveillance
try duck player
> https://youtu.be/MiUHjLxm3V0
PSA: the si (along with pp) parameter is used for tracking purposes:
?si=kEPSicC2WXYhcQ6L
Asianometry has half a dozen or so videos of you want some really deep dives on the tech and industry (with sources, since we’re on HN)
Okay this is weird.
> The key advancements in Monday's disclosure involved doubling the number of tin drops to about 100,000 every second, and shaping them into plasma using two smaller laser bursts, as opposed to today's machines that use a single shaping burst.
This is covered in that video. Did they let him leak their Q1 plans?
That has been covered before in other videos[0] that this is their roadmap to higher power, so I'm also not sure what they have announced now that wasn't previously announced.
From the first video I thought they had already shipped this, but it sounds like they were describing what their new model was.
This seems like a product with a very very long sales pipeline, so I wonder if they work on pre-orders with existing customers but announce delivery milestones only as they come?
One of those odd moments where a YouTube title looks like clickbait but is actually, factually correct.
+1 for this video, and the Branch education one. Well done to both teams.
As shown with that terrible speed of electricity video, Veritasium prefers "technically correct" over factually correct.
Highly recommend this video as well, he has a bunch more worth watching. https://youtu.be/rdlZ8KYVtPU?si=wgjkkNDSzuuS3lVK
Comment was deleted :(
> The company's researchers have found a way to boost the power of the EUV light source to 1,000 watts from 600 watts now.
> "We see a reasonably clear path toward 1,500 watts, and no fundamental reason why we couldn't get to 2,000 watts."
Why this is a big deal:
Right now the only way to make "bright" EUV (100-200 watts) is to spray fine drops of a metal in a stream, then target and blast each drop with a laser.
pretty wild way to make light.
And they’re now going to hit each drop three! times instead of two, and increase to 100,000 drops per second. Very hard to imagine.
If you wrote a science fiction novel around the idea that we make computing devices by blasting fine drops of tin in a vacuum with a laser exactly 3 times at exactly 100,000 drops per second, nobody would believe it. Truth is crazier than fiction.
What's even crazier is the technological pursuit of EUV and what a moonshot it was. Chip wars by chris miller chronicles it and it is absolutely crazier than sci fi.
> SAN DIEGO, California
> to help retain the Dutch company's edge over emerging U.S. and Chinese rivals
Great news, but what a strange attempt to equate the U.S. and China in this and build a narrative. Cymer was founded in San Diego.
Yeah it's an interesting angle in the article. The EUV light source technology is completely designed, developed, and manufactured by Cymer in California, which is a US company that ASML acquired in 2013. If export control agreements were not in place then ASML would have never been permitted to acquire Cymer. And if they are not enforced then the US would almost certainly require ASML to sell Cymer back to US ownership, TikTok-style.
The reality is that it's American technology that is used in ASML machines so I don't know why the article tries to frame it like it's a competition.
There is much more in an ASML machine, besides the UV source.
So the ASML machines combine technologies developed in various places, not only in USA, even if the UV source is indeed a critical component. While an ASML machine would not work without the UV source, it would also not work without many other critical optical and mechanical components.
If it were so easy to make a lithography machine when you have a UV source, Cymer would have remained an independent company or it would have been bought by a US company. Cymer has been bought by their only customer.
The same happens when you look at a PC, it is likely that it contains something essential that comes from USA, i.e. the CPU logic may be designed by AMD, but the manufacturing technology is designed in Taiwan, the memories may be designed and made in Korea, other chips may be designed and made in Taiwan, other components come from Japan, the PCB may have been designed in Taiwan, but actually made in China, and so on.
So yes, it has some important US technology in it, but there is a very long way from a CPU logic design to a physical computer and most of that rarely has anything to do with USA.
The same happens with an ASML machine.
While this is all true, I think it should be emphasized that the EUV source and supporting it with optics (Zeiss) and other machinery is the primary engineering effort around the development of lithographic systems.
The primary difference between this machine and its predecessor is the degree to which it has been optimized around the specific EUV light source.
Your take is also a bad one. No what asml builds is not American technology. Why asml succeeded is because they got tons of company’s and people to help them advance the technology of the chip industry. Yes it wouldn’t be possible without the Americans. But it would also not be possible without the Europeans, the Koreans, etc… what asml did was basically ask the technology leaders in each field to build their best product so that they can take their parts and assembly this awesome piece of technology.
My understanding based on reading Chip Wars was that Cymer only made the lasers (which are underneath the machine you see in photos). The rest of the mechanism with the tin droplets and cameras and mirrors to aim the lasers etc came from ASML and its other suppliers.
Which American rival would that be anyway? I have not heard of any.
xLight is the promising new US competitor to Cymer. Lots of funding from the US CHIPS And Science Act. Founded by Dept. of Energy engineers who formerly worked on large-scale X-Ray systems and particle accelerators.
Oh, alright. I was thinking of full lithography machines.
It wasn't just "permitted." The technology under discussion here was funded by the US the DOE and Intel and deliberately transferred by the US to ASML (and not, for example, a Japanese company or Samsung) as part of a soft power exercise.
It's crazy that Europeans keep citing ASML as a strong example of European innovation.
It is a strong example of European innovation though. It’s a multinational project. I wasn’t casting shade on ASML, I was just pointing out the wording of the article implies some sort of competition between ASML and the US that does not exist.
I think the Japanese are also working on potentially competing technology
Don't think Rapidus is. Nikon & Canon might be researching alternatives to EUV Lithography but how long is that going to take?
LSTC will contribute research (for 1nm) though it isn't clear if it'll be in partnership with ASML or as competition.
https://www.csis.org/analysis/japan-seeks-revitalize-its-sem...
The light power increase is even more impressive at 67%:
> The company's researchers have found a way to boost the power of the EUV light source to 1,000 watts from 600 watts now.
with more on the horizon:
> We see a reasonably clear path toward 1,500 watts, and no fundamental reason why we couldn't get to 2,000 watts.
So how small are individual components (e.g., transistors) nowadays? Presumably there's a lower limit: once you're a few atoms across, it seems that you can't go any smaller (?).
Gates are about 30-50 nm wide, even though they're called '3nm' for marketing reasons.
Metal pitch is 26nm. That means parallel wires can be placed 2 wavelengths apart with 13.5nm light.
As The Register used to call it, these clean-sounding process nodes (15nm, 5nm, 3nm etc) are "marchitecture." Marketing architecture. Reality is much messier.
Like free range chicken.
You only need to live in reasonable place for that phrase to have a proper meaning, across whole market from cheapest to most expensive.
What is 'proper meaning'? Only a single brand of egg in my supermarket is genuinely free range per the definition supplied by our agricultural scientists (CSIRO, Australia) - less than 1500 hens per hectare. "Free range" can mean anything up to 10,000 hens per hectare.
Comment was deleted :(
https://beterleven.dierenbescherming.nl/
They have a code in my country that is pretty simple from 1 to 3. The highest tier is 11 chickens per square meter apparently (I do not eat poultry myself).
This is about increasing output per machine via upgrades.
some gates are only 10-14 nm wide, thats about 50 silicon atoms!
I still think we should have gone with average gates per square mm as a new yardstick. It would also make sense to the Numbers Go Up people.
It’s going to be quite funny if they can go below 40nm in gate pitch size, because they’ll need to call it 0nm.
They are moving to angstroms, hence 18A for example.
Another marketing gag could be -1nm
And what hard drives and memory slots would those chips be able to use?
End result: the AI industry will get 50% more chips, the rest of us plebs will still be waiting for new GPUs to hit the market...
It's impressive to see that there still was so much room left to improve EUV, but I can't help but be royally pissed off that it will be a looooong time before we the people see any practical benefit of it.
Yes, and these days both AMD and Intel have announced that their next generation of desktop CPUs, Zen 6 and Nova Lake, which are expected to have very significant improvements over the current CPUs (including much more cores for both and AVX-512 for Intel), have been postponed for next year, instead of being launched this year, as originally planned.
This delay is presumably caused by both the lack of production capacity at TSMC, which is busy with AI, and by the fear that any launch of a new CPU would be crippled by the impossibility to buy DRAM and SSDs for new computers.
It's high time for governments to step in and do some Soviet style regulation. This AI shit is getting out of hand, it's not healthy for any economy when one group of actors has so much money that they can buy out entire classes of Things that are vital for any society to be at least somewhat affordable and widespread.
This is a steep increase of power to get out of a vacuum system that is highly sensitive to temperature changes.
[flagged]
Crafted by Rajat
Source Code