Extreme Ultraviolet Lithography

anymouse writes "The April issue of Scientific American has this article on chip lithography using soft x-rays (extreme UV). Strange bedfellows - Intel and AMD at Sandia." Discusses the research process and technology behind the next generation of chip production.

Comments

Re:But isn't there a limit

[stevelinton]

To nitpick, EUV is not actually a laser technology at all. The laser is just used as a heater to get really, really hot xenon gas, which then radiates thermally.

More seriously, the distinction between EUV and soft X-ray is putely a marketing one. This technique was called soft X-ray until X-rays got a bad name.

Re:It's a Good thing ...

[valdemar]

I can give this a shot. My knowledge is not complete in this subject, but I know a bit. First of all, yes, you are correct that SRAM and CPU's use a simmilar components and DRAM's use capacitors for each memory cell. Thats why DRAM has to be refreshed every so often, the capacitors lose there charge (They are really small). In order to store one bit in and SRAM cell you need to have a feedback circuit consisting of a couple gates. Basically you need a flip-flop.

Now here is where I dont know exact numbers. There are all kinds of tricks once you are playing with the silicon that make circuits designed at that gate level smaller than they would be if you actualy made each gate and connected them together. There is a great example of a 2-to-1 Mux that can be made in quite a bit less space by using a pretty cute trick.

Anyway, you need a couple transistors to make this flip-flop (or whatever the SRAM cell is) and some connections (lets ignore the connections for now). So, lets say you need 8 transistors (random guess, couple for the actual bit of state, and some to enable reading and writing the data). Current CPU's are up to tens of millions of transistors (the thunderbird has something like 37 million transistors). Thats only 4 megabits of SRAM, or 512KBytes. This number is not right, you also have to take into account the fact that its MUCH easier to route signals in a RAM chip. So it might be off by and order of magnitude or two. But even if it is thats still not much RAM.

I hope that helps, and I hope I havn't completetly messed something up. Well, I am sure someone more knowledgable can correct me where I am wrong, and will. :)

Re:But isn't there a limit

[WNight]

Most people now say "The farthest we can go with technology X is ..." instead of just saying there's an absolute limit.

And it's a reasonable thing to say. You won't get trace size under twice the wavelength of the light you use to etch it. That's really obvious.

So you either read that as there being a hard limit on what can be done, or as someone saying a new technology is required to go past those limits.

I haven't heard any good "To go beyond X is impossible." quotes in quite a while.

old news

[expunged]

This is pretty old news. We have had presentations about EUVL at the university here once or twice in the last year and there is quite a bit of literature dating back to the mid-90s (I did a paper on EUVL for a course in semiconductors).

It is nice to see it in Scientific American, but I think EUVL has been brought up in discussions of other NGLs here on /. The article does take a good broad perspective on the issues as they stand.

Intel has a paper on their website (if you can find it) that describes the process pretty straightforward as well (it might help the read to have a little bit of background).

Here is that and some other URLs:

http://www.llnl.gov/str/Sweeney.html

http://developer.intel.com/technology/itj/q31998/a rticles/art_4.htm

http://lithonet.eecs.berkeley.edu/network/backgrou nd.html

http://lasers.llnl.gov/IST/euvl.html

http://www.lbl.gov/Science-Articles/Research-Revie w/Highlights/1998/ALS_chips.html

http://chomsky.stanford.edu/~kevbert/neha_poster/s ld001.htm

http://www.cr.org/publications/MSM2000/html/W3202. html

http://www.google.com/search?client=googlet&q=EUV% 20lithography

-nicole

Interesting to see...

[Deluge]

...how soon they'll be able to capitalize on the EUV technique. NVIDIA is only now moving to .15 micron production, Athlons are being manufactured at .18, .13 is what everyone's aiming for but the process still seems to need refinement.

This certainly isn't the 1st article to preach the wonders of EUV lithography, and should have dropped that "coming soon!" tone of the previous articles hyping the new tech. Fact is, chipmakers still have a lot of work ahead of them perfecting the use of current lithography processes, and squeezing all they can from what they know to work.

Yes, research into future possibilities of chip manufacture is important, and Intel has been wise in not abandoning research in that area (duh) but they're not the heroes the article makes them out to be, they're only making sure they won't get left in the dust when current techniques become outdated. And they will not allow the consumer to benefit from that new research until they've milked the consumer for as much as they can selling them products created with current tech.

---

Re:But isn't there a limit

[Deluge]

Yes, I see the '40 atoms in width' line. The heat from each of the transistors would screw up the transistor beside it. That was the limit of the copper lines within a chip. Now they're saying they've overcome that limitation?

D'ya ever get the feeling that these researchers and scientists and engineers working on chip technologies all get together at their IEEE meetings or some such thing and conspire against the world to make themselves seem smarter than they are?

And I think they do this by agreeing to publicize some arbitrary limit, a roadblock to whatever they're doing, causing all sorts of worry among the general population about what'll happen when that roadblock is hit, and then when the roadblock is approached, "OMG OMG we figured out something new, this roadblock is no longer an issue, move on folks."

It happens so often nowadays that nobody even notices it. How famous are the quotes by well known individuals claiming things like impossibility of supersonic travel, impracticality of computers, even impossibility of flight itself? Now we hear "once we reach this point in our technological sophistication, we can go no further" every 5 minutes, only to have that claim disproven another 15 min later.

Ok, i'm done.

---

Re:You know you've been staring at the screen to l

[The_Messenger]

You mean Intel's International Trio of Gay Blue Men? Yeah, they're pretty fucked up. My question is, how much acid do you have to do before they go away?

--

Re:fp fp fp (Getting as offtopic as possible...)

[The_Messenger]

I can post de-obfuscations at +2. ;-p

--

But isn't there a limit

[HerrGlock]

Yes, I see the '40 atoms in width' line. The heat from each of the transistors would screw up the transistor beside it. That was the limit of the copper lines within a chip. Now they're saying they've overcome that limitation?

Printing with UV lasers, no matter how sexy this might be, seems to be safer than the x-ray technology they were using. I would much rather be in a lab with the requirement for full covering goggles than have to wear a lead lined jock strap.

DanH
Cav Pilot's Reference Page

Re:But isn't there a limit

[Alien54]

well, eventually moore's law will have to give out. As mentioned in the article:

Whether Intel's buy-it-when-you-need-it strategy can work more generally remains to be seen. The real test may come in 15 years or so if EUV or EPL gives out and some wholly new substitute for silicon chips is needed. A paradigm shift--using molecules of DNA, nanotubes, quantum dots or other exotic materials to execute computations--may determine whether the virtual-research model can succeed. "Intel did a magnificent job of picking up the technology, recognizing its worthiness and driving it home," Freeman says. "But they're not putting the same effort into asking the questions about what to do when you get to 100 angstroms [10 nanometers]." Maybe one of Moore's successors will have to lay down the law for quantum computing.

with the heavy investment etc onresources riding on the free lunch provided by Moore's law, it is going to be one heck of a financial collapse when Moore's law gives out. You thought the dotcom collapse was bad? just watch what happens when Moore's law gives out.

Thanks for the insider info...

[HerringFlavoredFowl]

Thanks for the info on the direction they are going after 157nm. Tells me which way to point our test equipment (Ashame our web mistress took our VACUUM UV products off the website).

And yes we do make test equipment that goes down to 13nm.

Excellent article, if you believe everything you read, they will completely skip 157nm and go straight to 13nm. I think 6 to 10 years is more realistic.

TastesLikeHerringFlavoredChicken

You know you've been staring at the screen to long

[AliasTheRoot]

...when you misread the title of this as Extreme Ultraviolent lithography.

Competitors? allies?

[popoutman]

I think that it would suit Intel to have some of the r&d costs spread around, and Intel cannot afford to have the ability to produce 'better' parts, without the 3rd party support in chipsets/ram etc. (looking to maybe 4-6 years in the future..). Intel would not invite the others in the industry unless it suited Intel somwhere along the line.

off-topic: Re:Competitors? allies?

[popoutman]

I think it would suit your lowly self to adopt a large pile of SPAM into your home, for just 40 cents a day you can give SPAM the warm, caring place in your inbox it deserves. Please, give generously, the marketing department thanks you. pio@skynet.ie ...the de-obfuscatah, because mailinglists need love too. Doesn't bother me at all my friend. Spam I get, spam I bin. Nice to know people like you exist, just proves how shallow the other end of the gene pool is

:-P

hehe. p

Stuff is already so small what's it all for?

[BroadbandBradley]

I can see setting my PDA down only to watch it get picked up by the wind and blown away. Of course more power is always nice, but at what point would it be cheaper/easier to just start stuffing multiple cpu's in a case and Multi-thread stuff? you gotta figure that this new technology is expensive and at some point having 10 2Ghz chips in a box might be more effective then buying a machine with one 20Ghz chip.
what's it all for?
see the article in the same issue on tele-immersion where a 3d world is created on each end for virtual meetings. It's great stuff but it takes racks of computers on each end to get about 2-3 frames per second. I guess we really do need faster computers yet....