by Rucker, Rudy
As well as having to be positioned to an accuracy of a tenth of a micron or better, the successive layers need to have a very specific thickness. Rather than being measured in microns, the thickness of the layers are best measured in nanometers, or billionths of a meter. Each layer is about ten nanometers thick. It’s all about fiddling with little details, to a mind-boggling degree.
The process takes as long as twelve weeks for a completed wafer’s worth of chips. It’s not so much a linear assembly line as it is a loop. Over and over, the wafers are baked, printed, etched and doped. At AMD, workers carry the boats of wafers up and down the corridor; while at Intel’s plant there is a miniature overhead monorail on which the boats move about automatically, like gondolas in a scale model of an amusement park ride.
At AMD, I visit the etching bay first. There are a series of sinks filled with different kinds of acid piped up from tanks located on the story below the fab. In the bad old days, you could recognize fab workers by the scars on their neck from splashes of acid, but now they have a small industrial robot arm to dip the chips. I’m happy to see the arm; this is confirms my science fictional notion that fabs will ultimately be places where robots reproduce themselves: robot obstetric wards.
The acid baths are for removing the photo-resist masks after the etching itself is done. The etching is typically done “dry”—that is, a fine dust of ions is whipped into a frenzy with powerful radio frequency signals to make a submicron sandblaster. The idea is to dig out parts of the chip so that metal conductors and metal-doped semiconductors can be patterned in to make up the wires and transistors of the integrated circuit which the chip is to become.
The real heart of a fab is the photolithography bay. Here the gel called photo-resist is sprayed onto the wafers, and then the wafers go into a stepper, which is the machine that projects the circuit diagrams onto the wafer’s chips.
The projector is called a stepper because it projects the same image a hundred or so times onto each wafer, moving the wafer in steps to receive each successive image. Steppers are the most expensive devices in a fab. The images projected by the steppers are found on transparencies called reticles. Reticles are based on circuit diagrams created by engineers using computer drafting techniques.
Once a wafer gets out of the stepper, a developer chemical removes the photo-resist that was exposed to light, leaving masks shaped like the dark regions of the reticle. This is a very efficient process, because although a reticle may have thousands of features on it, projecting its image onto the wafer puts all those features there at once.
The better the stepper, the smaller the images it can make. Smaller chips run faster, use less power and can be produced in larger batches—more chips per wafer. In order to handle very small feature sizes, steppers need to use light with very short wavelengths—the current ones use deep ultraviolet light, and to get much smaller, the steppers will have to start using X-rays.
The light is mellow yellow in the bay with the steppers, and there are the most people here. This is the heart of the temple. Some of the workers are debugging a problem with one of the machines that sprays on the photo-resist; one of them is lying on the grated floor with a laptop computer. It strikes me that in this world, the floors are not dirty.
There are a couple of men with an electron microscope looking at wafers. One of them is holding a handful of wafers, some of them cracked. “I guess those ones are no good?” I ask. The man looks at me oddly and finally grunts, “Yeah.” Seeing only my Visitor badge and not my expression, he thinks I’m an executive being sarcastic, but Don explains that I’m a journalist. The guy warms up then and has his co-worker show me some wafers under the electron microscope. There’s a nice clear image on a TV screen next to the microscope. It shows something like your usual image of a chip, but with lots of parts missing. This is just one or two layers’ worth.
“These things,” the man with the microscope says, pointing to some fat short rectangles, “we call these the hot-dog buns. And these other things,” he points to some longer thinner rectangles overlaid onto the fat short ones, “we call these the hot-dogs. We check if the hot-dogs are on the buns.”
We peek into a few more bays. One especially cute little industrial robot catches my eye. It’s jerky and articulated like a shore-feeding bird, folding its tail and pecking wafers out of their cartridges to slide them into some machine’s maw. It reminds me of the Disney cartoon of Alice in Wonderland, where Alice is lost in the woods near the Cheshire cat and a little bird that looks like a pencil with two legs comes running up to her.
Dan takes some pictures of me, and then we go out into the gowning room to take off our face masks, gloves, and Fibrotek suits. It feels very good to get out of the suit, I was getting hot. Also it’s great to stop breathing my own breath. It would be tough to spend twelve hours at a time in a fab. And for $24,000 a year! As a communist friend used to tell me in grad-school: the secret of capitalism is that the less they pay you, the harder you have to work.
Now we’re in our building suits again, and Dan wants to show me the sub fab, which fills the whole story below the fab. As we go out into the building hall, a security guard in a clean room suit runs up to us and asks our names. He writes our names on his glove; he’s too excited to get the spelling right. He doesn’t recognize Dan, and we’re both wearing Visitor tags and Dan is carrying a camera. Uh oh. While the guard hurries off to make a report, Dan hustles me down the stairs to the sub fab.
The sub fab is a techno dream. It holds all the machinery that supports the machines of the fab. The electrical generators are here, the plumbing, the tanks of acids, the filtering systems, the vacuum lines, the particle monitoring equipment—miles of wires and pipes and cables in an immaculate ten-thousand-particle-per-cubic-foot concrete room. This is the ultimate mad scientist’s lab. I’m enthralled.
Now here comes the clean room security guard again. “You have to come with me.” Dan wants to take some pictures first. “You have to come right away.” The clean room guard leads us out into a hall off the sub fab. Three unsmiling uniformed guards are there. Dan explains about his lost fab badge; they phone the pregowning room to go into Dan’s locker and check out his ID; finally they decide it’s okay and we’re back on our way.
“They thought maybe we were from Intel,” Dan says. “Someone who doesn’t know me saw us taking pictures in the clean room.”
When I’m finally out in the dirty real world again, I’m grateful and glad. It feels as if I’ve been in the underworld, a world where people are totally out of place. I don’t feel like turning on a computer again for several days. But I’m happy to have seen the central mystery, to have penetrated to the heart of the temple of the computing machine.
Two weeks later, Intel finally comes through with a fab tour for me as well. My guide here is Howard High, of Intel Corporate Communications. The fab layout is quite similar to AMD’s although Intel’s fab is much bigger—perhaps the size of a football field, and with high fifteen-foot ceilings to accommodate the wafer-boat carrying monorails overhead.
The vibes in the Intel fab seem more relaxed than at AMD. Intel is ahead, and AMD is trying to catch up. At Intel, for instance, I don’t have to exchange my clothes for a building suit, I’m allowed to just put the clean room bunny suit on over my clothes. Because of dust, I wasn’t allowed to use any paper on my AMD tour, but Intel issues me a spiral notebook of lint-free paper.
The more I learned about the fabs, the more I was amazed that they work. The intricacy of the system is reminiscent of the complexity of a biological process like photosynthesis. Nobody could have designed one of today’s fabs from scratch—these are giant industrial processes that have evolved, a step at a time, from earlier, simpler versions. There is a very real sense in which these processes are the synthetic biology in which planet Earth’s next great species may arise.
* * *
Note on “Robot Obstetric Wards”
Written in 1994.
&nbs
p; Appeared in Wired, November 1994.
This was the second of my Wired journalism runs. I had a lot of trouble setting up the visits to the fabs—at Intel they told me the last person they’d let in had been the vice-premier of China. But I persisted. It was amazing to get inside these secret temples of high-tech. I’d wanted to call this article “Fab!” but Wired’s title is probably better.
Goodbye Big Bang: Cosmologist Andrei Linde
Andrei Linde is a Moscow physicist who became a Stanford University physics professor in 1990. He lives there with his wife Renata Kallosh (also a Stanford physics professor, specializing in superstrings and supergravity), and his two sons Dmitri and Alex. He began formulating theories of the “self-reproducing chaotic inflationary universe” in 1983 as an improvement on the Big Bang model. He uses computer simulations for a lot of his research, and has recently suggested that your universe might be the result of a physicist-hacker’s experiment.
I went to interview him at his home for Wired magazine in the spring of 1995. Linde is an attractive, tidily dressed man, younger and more athletic-looking than I’d expected. He speaks with a thick Russian accent, and with a colorfully inverted syntax. His verbatim answers were sometimes bit cryptic—especially for non-physicists—so I padded a few of them a bit, mostly using materials from his published papers (such as Andrei Linde, “The Self-Reproducing Inflationary Universe,” Scientific American, November 1994).
RR: By now, most of us have gotten quite comfortable with the big bang model of the universe; the notion that the universe was born as a tiny energy-filled ball of space some billions of years ago, and that this ball of space has been expanding ever since. What’s wrong with this notion?
AL: There are a number of problems with the big bang theory; let me mention two that are of a physical nature and two that are of a philosophical nature.
If you work out the physical equations governing the big bang, they predict that a big bang universe will in fact be very small, even though we can see that our universe is large. One way to gauge the size of a universe is to talk about how many elementary particles it has in it—how many electrons, protons, neutrons, and so on. When I look out of my window, the matter I see is made up of perhaps ten-to-the-eighty-eighth elementary particles, but a typical theoretical big bang model has only about ten elementary particles in it! This is perhaps the most serious problem with the big bang model. It gives a false prediction about the size of the universe. For a number of years, this mathematical flaw in the big bang theory was not yet noticed.
A second physical problem with the big bang is that even if a big bang universe is of the proper size, there is no explanation of why the different regions of the universe resemble each other. In a big bang model, it could just as easily have happened that most of the matter ends up, say, in one half of the sky, but we can observe that in our universe, the density of distant galaxies is the same in every direction.
One of the philosophical problems with the big bang is this: What came before the big bang? How did everything appear from nothing?
Another somewhat philosophical problem with the big bang asks: Why does it happen that our universe worked out to be just the way it is; why, for instance, do we have three dimensions of space and one dimension of time?
The big bang theory offers no satisfactory answers to these questions, but we can begin to resolve the puzzles in the context of the theory of the self-reproducing, inflationary universe.
RR: What is the inflationary universe?
AL: There have been several versions of this theory. The first was proposed by the Soviet physicist Alexei Starobinsky, but it was rather complicated. Then a much simpler one was put forward by the physicist Alan Guth of MIT; we call his model “old inflation” now. Guth took the big bang model and added the idea that in the beginning the universe expanded very rapidly; faster even than the speed of light.
By having the universe expand so rapidly, you solve the problem of why it is so big, and you also solve the problem of why all the regions of the universe we presently can see resemble each other. The idea is that, thanks to inflation, the whole visible part of the universe was inflated from some very small and homogeneous region, and this is why we see large-scale similarities.
It turned out that Guth’s “old inflation” had theoretical difficulties. I invented a “new inflation” theory which worked so-so, and then I realized that we could have inflation without the assumption that the universe began in a hot and dense state. I dropped the idea of the big bang, but kept the idea of inflation. In my model, inflation can start anywhere. This concept is called “chaotic inflation.”
RR: What causes the inflation?
AL: There are things called “scalar fields.” These fields fill the universe, and show their presence by affecting the properties of elementary particles. You don’t notice a constant scalar field, any more than you notice a constant air pressure or a constant electric charge. When there are differences in air pressure, you get wind; when there are differences in electric charge, you get sparks; and when there are differences in the scalar field, you get an expansion of space.
Quantum mechanics implies that the scalar fields undergo unpredictable fluctuations. If there is a place where one particular scalar field happens to be larger, then here the universe will expand with a much larger speed, which makes so much space that we can safely live there.
RR: How big is the inflationary universe?
AL: The fluctuations which increase the speed of inflation can happen over and over. They make the universe self-reproducing; it reproduces itself in all its forms.
The standard big bang theory was a theory of a homogeneous universe, looking like one single bubble. But if we take into account quantum effects, the self-reproducing inflationary universe is a bubble producing new bubbles producing new bubbles.
This kind of repeatedly branching pattern is what mathematicians call a fractal. A fractal pattern is characterized by the property that the small bits of the pattern resemble the whole pattern. An oak tree, for example, is like a fractal in that a single branch of an oak resembles a scaled-down model of the entire tree. Another example of a fractal is a mountain range. If you chop off the top of a mountain and look at it closely, it resembles the whole mountain range; and a single rock on the mountain resembles a whole mountain in itself.
So we think of the self-reproducing inflationary universe as a fractal. The big bang is good as a description of each particular bubble but it cannot describe the growing fractal. There is no real reason for the fractal universe to stop growing; indeed, it is likely to keep growing and budding off new regions forever.
RR: How can I visualize the fractal self-reproducing inflationary universe?
AL: There are two kinds of pictures I like to use. In one I draw something that looks like lots of separate bubbles connected to each other where they touch. It looks a little like the linked flotation bladders on seaweed.
In the other kind of picture I use—and I’ve done several computer simulations of this image—I think of space as initially being like a flat sheet. Then I add a randomly fluctuating scalar field, and I represent the regions where the scalar field has a low value by valleys, and I represent the regions where the scalar field is large by peaks.
The peaks are the places where inflation takes place; at these places the universe will rapidly expand. I can’t show the inflation in my picture, but I can represent it by putting new, secondary peaks on top of the first peaks, third-level peaks on top of those peaks, and so on. It is like a mountain range.
What is a little hard to grasp is that the two pictures represent the same thing. The peaks in the one image correspond to the bubbles in the other image. A peak that rises on top of a peak is like a bubble that swells out from the side of a bubble.
RR: Can we go to the other bubbles of our fractal universe?
AL: Far in the future, our sky will start looking a lot different, as our stars start dying. And then we will
see into the different parts of universe, some parts with different laws of physics.
Can we use the energy in our bubble which has cooled off, can we fly to the other tips of the fractal, can we go there and live comfortably? The theory of such cosmic flights suggests that even if you travel at the speed of light, you lose so much time that when you get to another part of the universe, it will already be cold and empty there.
RR: You say that some of the different bubble-universes have different laws of physics—how does that work?
AL: We’ve talked about one scalar field that is responsible for the universe’s expansion. It seems that there may also be a second scalar field which makes different kinds of physics in different regions of the universe. There is one overall law of physics for the whole universe, but the scalar fields make for different realizations of this law. It is like water with many different phases. For those who live in water it is very essential that the water be a liquid and not a solid or a gas.
RR: What if I could somehow fly up to the edge of a region of the universe with different physics? How would it look?
AL: Between the different regions of the universe, there are boundaries called “domain walls.” There is a tendency of the domain walls to straighten up, and also to move one way or the other with a speed approaching the speed of light.
So first of all it would be very difficult for you to reach a domain wall if it is moving away from you. And if it is moving towards you, it will be very difficult to run from it. In fact, if a wall moves towards you at the speed of light, then you first see it only at the moment it hits you.
But we don’t need to worry too much; the typical estimates in these theories give you a distance from us to this next domain wall which is much much greater than ten billion light years, so we may live for now.
RR: Might we say that the regions with different physics are competing with each other?
AL: I think about the moving boundaries of the regions as perhaps like a Darwinian fitness. Should we discriminate and say those with greater volume are winners? There is a lot of place for losers as well, everything which can exist tends to have room for its existence in the self-reproducing inflationary universe. We can think of a Darwinian process without hate and killing, a process that produces all possible species.