Chips everywhere. “Not only in things, but also in ourselves,” says Alberto Sangiovanni Vincentelli (Milan, 1947), a professor at the University of California, Berkeley and responsible when it comes to revolutionizing electronic circuits. They are in cell phones, computers, cars, home appliances, toys and even in prosciutto, Iberian ham in Italian. Placing a slide next to a hog’s leg bone to measure the salinity and moisture of the ham—the two most important things when deciding whether it’s ready to eat—was one of his ideas. today is smart prosciutto It is a fact, as well as the sensors that soccer players wear to measure their performance or confirm their presence on the field. Both inventions passed through his office in Berkeley “about twenty years ago”. So is Elon Musk’s brain chip, something a colleague from AUC developed “a long time ago.”
Sangiovanni Vincentelli was ahead of his ideas and created the tools that made it possible to have these technologies today. He was awarded the BBVA Foundation Frontiers of Knowledge Award for his contribution to the design and improvement of chips found in electronic devices today. It has been a reference in the past 50 years for the transfer of knowledge between the academic and corporate worlds, facilitating the transformation of the chip trade around the planet.
The researcher speaks with EL PAÍS by video from a mountain town in western India, where he has traveled to attend a meeting of the directors of one of his 10 companies. Sangiovanni Vincentelli also founded the companies Cadence and Synopsis, references in the global electronics industry for developing the software used in each of today’s chips.
ask. How do you feel when your work is used around the world?
Answer. unbelievable. At some point, it comes naturally. You design a component with tools and it looks like the person who invented the hammer. Everyone uses them, but whoever built them didn’t think, at the time, that they were building every hammer in the world. What we do is help people design chips that go into every object today.
s. When you started your business 50 years ago, could you imagine this would happen?
R was found. Yes, of course. It wasn’t hard to imagine. It was clear that smaller, more powerful, and cheaper chips could be made. And so on and so on. You can imagine all kinds of applications. I mean, I couldn’t do it, but it was clear I was headed in that direction. In fact, one of the founders of Intel, Gordon Moore, posited Moore’s Law, which states that the number of chips on a substrate will double every two years. This has remained so far. It’s incredible, because it was 45 years ago. Many of us have the same idea.
s. Is it physically sustainable to maintain this parameter?
R was found. We say many times that we can’t do that anymore. I remember a good colleague of mine and responsible technologist, Professor James Mendel, when transistors were close to one micron. [equivalente a la millonésima parte de un metro] “That’s it,” he said in a speech, “we can’t take it anymore.” But we are now one nanometer away [la millonésima parte de un milímetro], which assumes the existence of a few atoms, one on top of the other. So not much can be done. Even devices are already so small that they no longer behave like a transistor and have become a random component, where you can’t really predict what’s going on.
We are approaching the physical limits, but not the end of microelectronics capability
s. What are the complications?
R was found. The first is the manufacturing process, how do you make something so small. The light no longer gets there and starts to use it electronic packages [haz de electrones] And the laser to be able to reach nanometers. But without this, it is not clear what can be done. The second is that due to its small size, it does not behave in a certain way. The last one, is it so expensive and is it really worth it? Currently, a new three-nanometer manufacturing line costs about five billion dollars. These three things reveal that we are approaching the physical limits, but not the end of microelectronics capability. We can do something called multi-chip packages.
s. Does that mean uniting multiple segments rather than having a single superpower?
R was found. This is what we’ve been doing since the 1940’s, when the transistor was invented. Place the components on top of the trivet and connect them together. On the other hand, making a single integrated circuit on a single chip is fun to perform, because it runs faster, consumes less power, and can be cheaper to manufacture. But when development is too expensive, you have to go back and use the old formula. Now, what you want to do is link to many abstract chips. Instead of being boxed, wrapped in plastic, and attached to the board, they are only used as they come off the manufacturing lines. It looks like a cake, but with a thousand layers. The distance between them is too short, so the performance is not good, so it is a compromise. Now they are being called Shiplettwhich is a funny name for multiple modules that we started thinking about 30 years ago.
s. Why hasn’t he gone this route since then?
R was found. There have been companies that have tried to do this, but they have all failed. The reason is that technology was advancing very quickly. By the time the package was multi-chip, there was already one chip that had it all. Even if one slide was bare, the other would have been better. But now we can’t squeeze more things, so we must restore the birth of technology.
s. Is machine learning overestimating the microprocessor scenario?
R was found. Yes exactly. Unfortunately, technology is advancing by leaps and bounds and some leaps are made on faith. Something looks good, it’s predictable and what it can do is extrapolate. Machine learning is just one of the chapters in artificial intelligence. Imagine there is a black box and you want to decode it. So an experiment is done: you put something in it, see what comes out, and then try to guess what’s in it. And here comes the difference between physics and machine learning, which in my opinion is just one approach to identification. What you did in the old days with a black box was try to figure out the physics inside it. Then I noticed this phenomenon and wondered what could explain it. This is a key point. Our minds, through modern mathematical models and experiments, try to explain why. The AI can’t do that, because it doesn’t know the mechanism behind the device behaving this way. But if you want to search the web with ChatGPT that’s fine, no problem there. If you lose something, no one will die. But if it’s a failure in self-driving, that’s another story, anyone can die.
s. How far can this intelligence go?
R was found. One of my colleagues started asking leading questions and at one point ChatPGT said this is crap. You can cheat. In terms of the intuitive way of using the internet, I sometimes get irritated because if I’m in Singapore or India and I’m looking for a place I can see [un partido de] Milan vs Tottenham, I get all the local channels. But this does not interest me, because I want to know the Italian channel. Even if I say “in Italy”, it still gives me all the channels from India. However, if you tell ChatGPT “I want to know which channels show AC Milan”, it will do so without any problem. Take the frustration out of spending time looking for something that could have been done smarter. And you can implement it interactively, I would love something like Alexa. Sometimes it does a great job and ChatGPT is definitely much more powerful than Alexa because it’s a bigger model and therefore can squeeze more knowledge into my past.
s. What would you like to see in the coming years regarding this type of technology?
R was found. The main point is to understand what your limits are. Instead of saying it’s all good, it’s asking what’s the downside to the technology we’re developing. There is always a downside, but people just don’t get it out there. To some extent I try to do everything I can, it would be good for everyone to study the pros and cons. About machine learning or ChatGPT, they are excellent. But you have to think about the downside, what can be guaranteed with this technology, where it is best used, and what it should not be used for. Think of CRISPR-Cas9 technology, which allows us to modify our genes in a precise way. It can eradicate genetic diseases, but it can become like the Nazis, everyone is born with blue eyes, you can do that now by the way. How do we make sure something like this doesn’t happen?
You can follow country technology in Facebook And Twitter Or sign up here to receive The weekly newsletter.
Subscribe to continue reading
Read without limits