[ad_1]
Transcript
Stephen Cass: Hiya and welcome to Fixing the Future, an IEEE Spectrum podcast, the place we take a look at concrete options to some huge issues. I’m your host, Stephen Cass, a senior editor at IEEE Spectrum. And earlier than we begin, I simply wish to inform you which you can get the most recent protection from a few of Spectrum‘s most necessary beats, together with AI, climate change, aynd robotics by signing up for one among our free newsletters. Simply go to spectrum.ieee.org/newsletters to subscribe. Right this moment, we’re going to be speaking about making tiny issues even tinier in order that we will cram ever extra computing energy onto silicon chips. And to try this, I’m speaking with one other Spectrumite, senior editor, Sam Moore, who covers a semiconductor beat for us like a field effect transistor masking a depletion layer. Sam, welcome to the present.
Samuel Ok. Moore: Thanks, Stephen. Nice to be right here.
Cass: So we’d typically speak about Moore’s law, no relation, on this present, and the present state of it and the way we at all times appear to be speaking about it’s the top of Moore’s regulation. And but, it retains going. So are you able to discuss a little bit about what the present state is and these new concepts for pushing that boat additional down the stream?
Moore: Certain factor. Sure, Moore’s regulation is slowing down. That’s a form of definitive reality. It’s getting tougher and dearer to make extra transistors on a given space of silicon. However it’s persevering with, and there’s plenty of effort to make that occur. Proper now, we’re form of across the 200 million transistors per sq. millimeter vary, and going to proceed to maintain attempting to make that smaller. You hear quite a bit about form of five-nanometer node, three-nanometer node, and stuff like that. You wish to take into account that these names truly don’t have anything to do with the scale of stuff on issues. The five-nanometer node— sorry, the five-nanometer node chips usually have form of their smallest distance between the wires is 20 to 25 nanometers. So it’s all only a title. However they will proceed to call new issues and make new processes and make issues even smaller.
Cass: So what do these names relate to in any respect then?
Moore: It’s historic. There was a time after they truly meant one thing. That measurement, it’s known as the metallic pitch. Principally, the space between two wires used to really be what they named issues after, excuse me. However that sort of broke down within the late ‘90s or so. And so ever since then, it’s simply been form of a reputation.
Cass: And what’s it primarily based on?
Moore: Sorry. So what’s it primarily based on now?
Cass: Yeah.
Moore: Oh, nicely, they saved chopping that distance in half and in half once more. After which they only sort of continued utilizing that form of division for the title, although it wasn’t truly associated to the scale of issues on the transistor.
Cass: It’s identical to a common, “We’re getting smaller sort of—”
Moore: Yeah, it’s simply sort of advertising and marketing. And there’s so few firms who can truly do these actually, actually cutting-edge chips, so they may name them something. And now they’re. I imply, TSMC now calls its course of N5. Intel goes to be— their subsequent technology or the technology after that will probably be like A20. I believe the A’s for Angstrom.
Cass: So for you, the actual measure, although, is that one you mentioned earlier, which is what number of million transistors you’re going to get per sq. centimeter sq..
Moore: Yeah. That’s actually what issues. It’s simply what number of you possibly can pack in.
Cass: So in the latest situation of Spectrum, we had a implausible characteristic that you simply edited called the Moore’s Law Machine about a few of these efforts to form of maintain Moore’s regulation going with this fantastically elaborate gadget. And this was written by Jan van Schoot. So maybe you may take me by this expertise, which is known as excessive ultraviolet, which appears like a soda I would purchase, however yeah.
Moore: Sorry. Effectively, they may have known as it comfortable X-rays. However excessive ultraviolet is certainly cooler. So excessive ultraviolet lithography is the way you make the— it’s the primary machine concerned within the newest two generations of cutting-edge semiconductor chips. So with out it, you wouldn’t have your iPhone 12 by 15, I imagine, the NVIDIA H100, that GPU that everyone’s attempting to get their fingers on to do their AI. You wouldn’t have a minimum of one of many prime 10 supercomputers and possibly not one of the subsequent technology of them. That is the crucial machine, and it’s made by one company in the Netherlands. And it’s fantastically sophisticated. Let me form of inform you first what it does, after which I’m simply going to offer you some bizarre superlatives about it. What lithography is, is form of, principally, you’ve received a sample that you simply wish to venture onto the chip that can ultimately make all of the circuits and transistors and issues like that. And with excessive ultraviolet lithography, you might be utilizing a wavelength of sunshine that’s solely 13 and a half nanometers lengthy. This can be a enormous leap from what was used within the earlier technology, which was 193 nanometers, which was known as deep ultraviolet. It’s an unlimited leap. It took greater than 20 years of R&D to really get to a machine that works and that’s vaguely reasonably priced. And after I say vaguely reasonably priced, I actually imply vaguely reasonably priced. Every machine is greater than $100 million. It’s received 100,000 or extra elements in it. It consumes a megawatt of electrical energy simply so it could actually ship a few hundred watts of this excessive ultraviolet mild onto the silicon wafer.
The factor weighs like 180 tons. I imply, it’s large. The present technology is like trailer measurement. I noticed one among them being put collectively at a fab in upstate New York about 5 years in the past. And it’s so huge everyone seems to be like an Oompa Loompa. I imply, very. If the Oompa Loompa’s have been the very best on the earth at what they did and the chocolate manufacturing facility value a billion {dollars}. It’s simply a tremendous machine. And the subsequent model of this machine, which is form of what we’re going to speak about right this moment, is greater than a 3rd bigger than right this moment’s. So it’s simply large, sophisticated, tremendous costly, laborious to get your hand on. And I wish to inform you how they really made it higher.
Cass: So inside this large trailer machine, there’s some actually loopy elements, together with how they really make these comfortable X-ray/excessive ultraviolet blink beams. And it includes molten metallic and a carbon dioxide laser.
Moore: Yeah, it’s the most bananas course of you possibly can sort of consider. So that you’ve received a vacuum chamber and a little bit— I don’t know what to name it, nevertheless it’s spitting tiny molten drops of tin. And so they shoot throughout the vacuum chamber, and then you definately hit that tin together with your 40 kilowatt laser. You blast it right into a plasma. After which this plasma glows in every kind of implausible colours. However the optics accumulate the 13 and a half nanometers that you simply truly wish to use and venture it into the center of the machine itself.
Cass: This actually does sound very Wonka-ish.
Moore: Yeah, it seems like there ought to be a neater method to do that, however apparently, there isn’t.
Cass: So these items must be vacuum-sealed as a result of this mild will get absorbed by the air. And what are a number of the different— why is that this machine so huge? As a result of it looks like you’ve received like a little bit tiny sample. You’ve received little tiny chips, okay? These little tiny droplets. Why is it so huge?
Moore: Effectively, plenty of it’s truly the chamber containing the optics, that are simply insanely exact. The mirrors are fantastically costly. These aren’t simply strange mirrors. They’re a number of layers of alternating unique stuff as a way to get this sort of mild to mirror in the proper course with any effectivity as a result of UV is absorbed by tons and tons of stuff, together with air. And so plenty of it’s simply getting the sunshine the place it must go effectively with out disturbing any of the patterns that you simply truly wish to venture. After which there’s plenty of it that’s additionally simply dealing with the wafers and dealing with the masks, which after they’re truly form of in place, they received to be dealt with at nanometer precision. So these are extremely superb transferring machines.
Cass: So one of many huge challenges I believe that Schoot talked about within the article was, “Sure, you possibly can set this up, and you may get the lasers going, the machine’s going, however your throughput goes to be very, very, very low, uneconomical for the scale of the machine until you strive a few different methods on prime of the, once more, molten droplets being blasted by a really highly effective carbon dioxide laser,” which I do know I’m hung up on, however inform me a little bit bit extra about these different methods.
Moore: Certain. Okay. In order you’re a little hung up on— one of many greatest issues they needed to clear up simply to get to the primary technology was to make the tin explode in additional brilliance in order that you may get simply that couple hundred watts of sunshine as a result of, the dimmer it’s, the longer it’s important to expose the wafer. And so it’s all about throughput. That drawback is principally solved. However as a way to proceed Moore’s regulation, you don’t simply need this. You’ve received this mild, however you wish to truly maintain making smaller and smaller options with this mild. To try this, there’s three knobs which you can flip. One was the large knob of adjusting the wavelength, which is smart. Smaller mild, higher decision, completely simple. Two different knobs. One is sort of tough to clarify. It’s a bunch of optical methods that you are able to do, which could embrace as a lot as projecting two patterns serially to get one sample on the finish or simply making issues look bizarre in order that they appear much less bizarre after they get to the silicon wafer. And I can discuss extra about that later.
However the knob that they’re turning with this latest machine that’s presently being inbuilt Belgium proper now could be to extend the numerical aperture. That’s form of the angle of the sunshine which you can function inside the optical system. Traditionally, they’ve turned all three knobs. Numerical aperture is one that really provides you a very good return, traditionally. And they also actually wished to do that. They’re proper round 13.5 nanometer decision now, but when they wish to dip down beneath, they’re going to must do a excessive numerical aperture, excessive ultraviolet lithography. This causes a cascade of issues once you’re designing the system. This machine is already fantastically sophisticated. However as with every optics, you tweak one thing right here, it’s going to have another impact in a while. So let me undergo the cascade of the issues that they needed to clear up as a way to make excessive NA excessive ultraviolet lithography.
Okay. So first, you need the numerical aperture enhance on the wafer itself. That’s the place you’ll get the decision. However which means you even have to extend it on the masks. Now, the masks is the place you retailer the sample. So that you bounce the sunshine off the masks and goes by the optics, after which it lands on the silicon wafer, and that’s your level. Right here’s the factor. So you bought to bounce the sunshine onto the wafer— sorry, onto the masks after which off of the masks. And right here’s the place you’ve got form of a Ghostbusters second. These two streams can’t cross. You suppose unhealthy issues will occur. It would be bad, I believe, is the road, proper? So you possibly can’t cross the streams. So which means it’s important to angle them away from one another, okay? However you possibly can’t angle them away from one another an excessive amount of as a result of these actually specialised mirrors solely work as much as about 11 levels. And as a way to simply— should you simply wished to do that with none changes, you’d want 18 levels. So then they’re like, “Okay, nicely, now we’ve received to resolve. We’re going to want this angle. There’s no getting round it.” So the way in which they solved it was by growing the demagnification, which I do know sounds sort of growing a D, however principally, it’s shrinking stuff down quite a bit. They elevated it by like eightfold or one thing like that.
So like, “Okay, hey, drawback solved.” However not likely as a result of now your sample on the wafer is absolutely small. It’s like a postage stamp as an alternative of— not your sample, sorry. The quantity of wafer which you can venture onto all of sudden, tremendous small. It’s like a postage stamp, and that signifies that it’s important to do extra postage stamps, which signifies that you’re—
Cass: As a result of these wafers are massive.
Moore: Yeah, they’re 300 millimeters throughout and so dinner plate measurement or so. And so should you can solely perform a little bit at a time, it’s going to take you longer to do a wafer. After which it principally turns into so costly, it’s not even price it. In order that they needed to clear up that drawback by doing one thing sort of bizarre. It was form of like sort of funhouse mirror results. Principally, they elevated the demagnification in just one course. In order that they got here up with these specialised mirrors that sort of would stretch issues out [laughter] and shrink them. And it had bizarre results. I imply, you truly must make the masks stretched out. So the sample that’s on the masks is sort of this funhouse mirror model of what you need on the wafer. However amazingly, that really does it. You continue to wind up with a little bit bit smaller than you’d like of a projection onto the wafer, nevertheless it’s acceptable so long as you enhance the acceleration of how briskly issues are transferring by the machine. So what? 5 issues to resolve?
Cass: Yeah, I believe so, yeah. Yeah. However this impact, it sounds a little bit bit like within the previous days earlier than all of us had widescreen, flat-screen TVs, generally once you have been attempting to be exhibiting like a cinema film on a TV and all of a sudden, the side ratio would get actually bizarre and distorted as a result of they have been having to squeeze in on one axis to make all of it match, particularly when the credit would roll and would get all distorted in a single course. And that sort of jogs my memory of that. However these are huge machines. And also you’ve carried out another reporting, although, on a number of the form of the negative effects of dealing with these huge machines, which is easy methods to function them in a extra form of environmentally pleasant method. And that was the work of this firm known as, I believe, Edwards in England. So are you able to inform me a little bit bit about that?
Moore: Yeah. So that you keep in mind how I mentioned every little thing has to occur in a vacuum on the within of this? Sort of form of not. There’s a very small form of circulation of— I imply, sorry. It looks like a small circulation of hydrogen, nevertheless it’s a very huge machine, so it’s truly 600 liters per minute. [laughter] However this hydrogen is there for a few causes. Every thing in there may be tremendous delicate. You don’t need something to get on the mirrors or on the masks or something like that. However you might be blasting molten tin in a chamber, and you’ve got different chemical compounds which can be concerned in chip making and stuff like that, and it’s essential sort of sweep them away. And in order that’s what this hydrogen is for. And also you suppose, “Oh, hey, hydrogen. That’s inexperienced.” Not but, truly. Most hydrogen is definitely not made in any inexperienced course of. It’s truly made by a chemical response between water and methane, so not nice. And 600 commonplace liters per minute is definitely sort of quite a bit. What they’re doing with it presently is they only burn it since you simply get water, and all the nasties that it’s picked up simply sort of falls out. And that’s—
Cass: Sort of a smoke stack on the facet with a flame on prime, [laughter] and it’s much more Willy Wonka. However yeah, okay. In order that they’re simply burning off the hydrogen.
Moore: Proper. However that’s tremendous wasteful. So what Edwards labored on was a system that may recycle the hydrogen. It’s truly fairly cool. It’s like a reverse gasoline cell sort of. The used hydrogen and the icky elements that it’s picked up alongside the way in which. Principally, undergo one facet, get ionized. Then an electrical area form of forces these protons by a proton change membrane. They arrive out of a membrane, excuse me. They arrive out the opposite facet, recombine with electrons. You get pure hydrogen to return into your course of. And all of the terrible stuff stays on that to the opposite facet. Yeah. So it really works fairly good. They set one up at a nanoscience analysis group known as IMEC, which is sort of a key European analysis home. And it managed to recycle 70 to 80 % of the hydrogen of their EUV machine. So now they only must persuade the large chip makers to undertake it as nicely.
Cass: So with all of those applied sciences— and I wish to flip to form of a competing expertise in a second, however for these applied sciences, how lengthy do you suppose it’ll be earlier than we see chips in our smartphones and our computer systems made with this new expertise?
Moore: Proper. With a excessive NA EUV, issues which can be made in 2025 will most likely begin to— a minimum of the chips themselves will probably be made in 2025. It takes months after that for them to be in programs, however that’s most likely in time for one more NVIDIA GPU. So AI is driving plenty of the demand for notably this most leading edge. And so I’d anticipate to see it in form of the technology of AI chips which can be form of made in 2025, ‘26. Additionally, Apple is at all times on the leading edge. They at all times need the latest chip manufacturing methods. So no matter iPhone [laughter] comes out within the latter half of the century will nearly definitely contain this.
Cass: And that’s truly an ideal segue since you talked about NVIDIA there. And NVIDIA is enabling one other strategy to squeezing issues down and conserving Moore’s regulation transferring alongside, which is inverse lithography. So are you able to inform me a little bit bit about that? And why the truth that it’s an AI firm works out for nicely for them? As a result of they’re a chip maker who occurs to make just like the AI chips.
Moore: Proper. So let me form of give us a little bit extra context because you began. So NVIDIA, truly, they design essentially the most in-demand AI chip on the earth. All people needs their fingers on an H100, which is simply the present technology. The producer of that chip, although, NVIDIA designs it. It’s manufactured by TSMC, which frankly, sort of dominates the world of essentially the most cutting-edge chips proper now. In order that they work carefully collectively now as a result of NVIDIA might be one among their greatest clients. So should you can sort of return for a second, keep in mind I advised you about these three knobs you possibly can flip to make lithography higher, to make your precision and your decision higher. A type of knobs was this bizarre one known as K1. It was form of the method stuff that you are able to do. So what NVIDIA has carried out is it’s made a type of course of stuff that you are able to do a lot simpler to compute. It’s a way known as inverse lithography. And right here’s the factor. You may suppose that should you wished to venture, say, like a plus signal, one thing that was formed like a plus signal onto a silicon wafer, in your masks, you’d put a plus, and then you definately’d get a smaller plus when it received to the silicon. And never the case. There’s sufficient distortions and different simply stuff that it’s important to fear about optically once you’re coping with this sort of operation beneath the wavelength of sunshine that you simply’re utilizing, that it’s important to do issues like add little form of canine ears on the finish of the plus signal to make it appear to be a plus when it will get there. These issues have needed to be progressively extra sophisticated as we’ve sort of pushed Moore’s regulation to its limits.
So now that plus signal would truly form of appear to be— should you put it in a kaleidoscope and sort of turned it, it’s simply this large bizarre stuff that it’s important to placed on the masks as a way to get your easy plus signal on the wafer. Now, these methods are literally sort of actually laborious to do computationally. So it’s the concept that like, “Okay, if I would like this plus on the wafer, what do I’ve to have on the masks?” And it’s so computationally tough that we’re speaking like weeks of simply— we’ve received an enormous pc, and it’s going to simply sit there for a few weeks and take a look at to determine what that form ought to be. What NVIDIA has carried out is it’s provide you with a system that turned that two weeks into an in a single day job. And the factor that was that it was once— it was once a job for CPUs. My guess is it was instructionally sophisticated sufficient that it was not form of inherently of the parallel nature that GPUs have been on. So NVIDIA did plenty of work and got here up with algorithms which can be simply completely fitted to a GPU. And so, principally, they did within the work— sorry, what would have taken 40,000 CPUs they did with 500 GPUs and two weeks versus in a single day, which is definitely— that’s enormous. That eliminates an enormous bottleneck in getting your chip to market, for one factor. It permits you to use this actually computationally costly approach in additional locations slightly than reserving it for the spots of the chip that have been simply actually tough.
And from the attitude of an environmental profit, it’s 5 megawatts of energy within the computing system versus 35 megawatts, which isn’t insubstantial. So yeah, this can be a factor that they— this computational lithography system, it’s known as cuLITHO. They launched it, I believe, within the early summer time or late spring, they usually’ve received Synopsis, one of many digital design automation firms purchased in. TSMC has been engaged on it with them. And naturally, ASML, which makes the EUV machine in query. And all the opposite lithography machine tools they’re signed on as nicely. So it ought to actually be making a distinction each environmentally and by way of getting chips carried out sooner.
Cass: So simply to wrap up, we’ve been speaking about plenty of applied sciences which can be truly very near being deployed. Is there something you’re seeing within the lab that’s additional out that may assist us like within the 2030s, principally?
Moore: There’s no clear reply as to if there’ll be form of one other wavelength of sunshine that we use, and it appears sort of unlikely, truly. Even 13.5 nanometers just isn’t that many atoms of fabric once you get all the way down to it. So our skill to form of shrink issues down in two dimensions, it truly is getting in direction of the top. And so transistor structure is beginning to go 3D. Or slightly, within the lab, it’s beginning to go 3D. However this looks like the trail that everybody has chosen. So now there’s a brand new sort of transistor. I imagine Samsung began utilizing it in manufacturing final 12 months, perhaps TSMC this 12 months. I may need these mistaken, however they’re each nicely into this new construction. It’s called a nano sheet. And Intel is transferring to it the top of 2024. And the factor concerning the nano sheet is that it’s form of conducive to creating a second transistor proper on prime of it. So as an alternative of attempting to squeeze issues collectively in two dimensions, we’re going to begin including layers. Along with simply on the transistor degree making it 3D, we’ve already received numerous work happening proper now and numerous manufacturing chips that contain 3D packaging, which is simply taking one chip and stacking it on prime of one other as a way to sort of make a superchip. And that’s occurring now in manufacturing chips. So yeah, the long run is three-dimensional.
Cass: Effectively, that was implausible, Sam. Thanks a lot for speaking with us right this moment.
Moore: It was a pleasure, Stephen, as at all times.
Cass: So right this moment we have been speaking with Sam Moore, senior editor at IEEE Spectrum about excessive ultraviolet and different applied sciences to maintain transistors getting ever smaller on pc chips. For IEEE Spectrum, I’m Stephen Cass, and I hope you see us subsequent time on Fixing the Future.
[ad_2]
