Semiconductors Are A Human Right

Dr. Naveed Sherwani is a serial entrepreneur and technical leader in the world of semiconductors. As a professor at Western Michigan University, he wrote several books and hundreds of articles about chip design and manufacturing. He was the founder of Open Silicon, PeerNova, and was the CEO of SiFive. Currently he’s the Chairman of the Silicon Federation, which aims to democratize chip design.

Dr. Sherwani’s Linkedin - https://www.linkedin.com/in/naveed-sherwani-2a789/

Timestamps:

2:11 - An explanation of the semiconductor shortage

5:37 - Roadmap of chip design to final product

9:25 - Semiconductors are a human right

11:41 - Semiconductors are the new oil

17:10 - How do we get semiconductors into the poorest countries?

20:08 - Silicon Federation (https://www.linkedin.com/company/siliconfederation/about/)

21:59 - How do you incentivize corporations to open source semiconductor technology?

24:49 - Will a new era of hardware lead to a Cambrian explosion in startups?

27:12 - The next two decades of chip development

30:25 - Getting young people excited about hardware

38:27 - Is China ahead of the US in artificial intelligence?

43:11 - To what extent is ML a hardware problem?

45:13 - Building a global education infrastructure 

50:00 - How do you democratize education opportunities for the world?

53:19 - In your opinion, what is the best piece of software ever built?

Transcript

0:00

I believe in order to access the internet, you need some kind of Silicon laptop, cell phone infrastructure, mobile infrastructure, some kind of. You know, wireless infrastructure, some kind of servers, some kind of cloud, all of that is built on a single technology today. Semi-conductors so if you believe that internet is a basic human, right? I would say semiconductor by definition is also a basic human, right? Because if you deny semiconductors and there's value, we believe everybody should have internet. Yet this nation or that nation where we will not let them have any semi-conductors. Well, if that is the case, then you're automatically saying that you're not let them have any internet. Dr. Sherwani. Thank you so much for joining us. It is my pleasure. Thank you for having me. Let's get right into it. So it's a really fascinating time in tech right now. And one of the hot topics is the ongoing shortage of semiconductor chips. I think in the last month we heard that some of the car manufacturers were going to shut down plants anywhere from a few days to a week. So it's really affecting real-world circumstances. So as a leader in the industry, what do you make of all of this? Well there are several reasons why we have shortages in that particular area of semiconductor segment and some have to do with the demand. Suddenly the electric vehicle demand has gone up and which is, you know a good thing we all think is a great thing, but I think more than that the. The car manufacturers typically are very used to buying commodities in bulk. They buy things like steel, rubber, other things, and many of those things have come complex supply chains, but not as complex as a semi-conductor supply chain. So. When these same commodity managers try to use similar kinds of techniques to get semi-conductors into their supply chains. I think they realize that the ordering system is much different, much longer planning is much difficult, much harder. So I think this industry is just beginning to learn. How to order very complex parts like semi-conductors I think in time it will get better, but that's one reason the supply chain complexity. Second is the increase of. The demand for Evie. The third of course, is that because of a trade war, many large companies are holding. They're trying to hold. They don't know which parts will be available, which parts will not be available. So what they are trying to do, they're trying to order extra, don't know where the water will end up. So they also created disruptions in the supply chain by. Forcing factories to build parts they want as a result, that capacity was taken away from other parts. So that is another disruption of the supply chain. So there's, and this is for, so this is but it sounds like these problems are for more than just cars. It sounds like they're for many, many products. Yes. But the, the many products like cell phone makers and laptop makers, and several makers are used to the complex supply chain of semiconductors. The difference here is the car makers may not be. And, but, but this is not something very complex they will learn. And then of course the trade war is not hopefully forever. And so disruptions caused by that will hopefully go away. So I don't. Well, nobody in semiconductor business thinks this is a long-term problem. We think this is a short-term problem and it will be, you know, within a year, a year and a half, we should be able to get on the other side of the supply problem. So a lot of the news was covering this as a problem caused by COVID, you know, car makers can't get enough chips to make most of their cars. So they'll have to focus on two things yet. Another casualty of COVID. Would this have happened regardless of COVID. Yeah, I don't think COVID has much to do with it. COVID has impacted on everything. There's no doubt about that because supply chains are disrupted all over the world, but semiconductor supply chains, I don't think factories are reporting that their output has dropped off. I don't believe that is the case. So I think the other reasons that we discussed are more relevant than the ones that we talked about. So maybe this is a good time to take a step back and talk about what the value stream is for going from chip design to ending up in a final product. That is how does a chip go from a glimmer and tip Cook's eye to ending up in the latest iPhone? Well, I, I think it is very, very complex system. When it started in 1940s or thirties probably it was very simple, but I think in last 30 years that the kind of semi-conductors we make today, the kind of chips we make today from the concept show of concept of a chip through his design phases. Through its manufacturing testing deployment of the systems that go along with the software development, into a things like a cell phone. We estimate that it goes through our thousand steps, go through about 50 countries, cross 70 borders. And, and most people are not used to most countries and most nations, most planners are just not used to this kind of a very complex supply chain, which spans the entire globe. So that's why people talk about, well, we will ban this country and we will do this. And we shut down this factory. All of that is quite meaningless when you are looking at some things which do spend or you know so many countries and so many steps, very complex steps, I think I'll give you an example. The wafer we use on which a chip is manufactured. It is it is a piece of Silicon and, and we make a cylinder cylindrical part of Silicon, and then we slice it horizontally. And the discs. And many of you must have seen these discreet called the wafers, these disks. Have to be forsaken. And when we say pure Silicon, this is what it means. It has to be 99.99999. There are five nines after the decimal point, pure Silicon, the total difference. In other words, if you look at the planet hop plane, that before is sometimes it is required that it should not be. The the entire mountain, the height of a mountain house, high height off of any bump should not be more than six items and height of any Valley depth of a Valley cannot be more than six items. So the total, from the depth of the Valley to the height of the maximum mountain on that Veeva, which is 99.999, 9% pure. It's six atoms. Now how many people, how many human beings on this planet earth comprehend? This statement I just made the measuring system for these Heights is itself, is incomprehensible. It's very difficult to even accurately measure and say, testing these things and figuring out how are they really, that pill is also a science of its own dimensions. So, so I. Statement when somebody asks me what complexity of semiconductor starting somewhere, ending up in something like a cell phone. It's a very long journey. It's very complex. It has taken us 60 years to get here and it took several different complex endeavors of human excellence from material science to software. To electrical engineering, to chemical engineering, mechanical engineering, process engineering, operations support, artificial intelligence. I can go on there about 30 different fields in here, which had to collaborate for the last 60 years to, to where we are. So so that's, I just want, I want to encourage your viewers to understand the complexity and appreciate the complexity because their views. And our opinions matter because in order to have an opinion on something, you have to understand the complexity of what you're talking with. Sure. And it really is a global effort, but even knowing this and knowing that it goes through 50 countries, is there a possibility for one country or a small group of countries to vertically integrate this manufacturing process and potentially monopolize chip makers? Well, anything can be when applets. It's not a thing that we cannot do it do. We'll obviously too, but I think the question would be is that what would be the cost of that monopoly? Right. So there are some industries which have monopolized, right? I mean, fighter manufacturing, right? I mean, There are a few countries in the world which make fighters. I believe they make fighters, but all by themselves, everything they do about making those, you know, four fifth generation fighters is built in those countries. They make a lot of money selling those fighters, but let's say if we had a civilian global development of those fighters, probably take 10,$20 million to make a fighter. These guys make them$400 million, sell them for 200,$300 million. So I think that's what happens when you try to limit it and do that. Of course, that's doable. The R and D becomes very expensive. The vertical chain becomes very expensive right now. What we are doing is that we are doing everywhere, things which are cheapest possible, and which is the best location so that we can get a cheap cell phone in the hands of poor people in the work. Well, if you do novelization, then rich people will definitely have cell phones. These people will definitely have laptops. Poor people will not have any laptops and cell phones. And that's what global disruption could mean. It means division the vets between the. Poor and the rich gets wider and it could be at individual level. It could be at national level, it could be at the company level. And so, so that's the reason, many of us believe that semiconductors have become so essential for life. Now they should, we treat it as like internet. Everybody believes that internet should be available to all humans. Similarly, we think semiconductors are something that should be freely available to everybody. We're not saying the cost. But access to that should be to all nations so that they can, you know, use them for the betterment of that nation. So you've previously stated that you feel that semi-conductors are the new oil. Can you explain what you mean by that? Yeah. So when not just me many people say that. I think what we mean by that is that when a resources scarce controlled by a few nations are people. Then they have an ability to utilize that, to make profit from it, and also deny that resource to people who they don't like or who they disagree with on some issues that can create disruptions, which oil did. So oil has a history of, you know 70, 80, 90 years of disruption of all kinds when the countries which have oil. Countries. We had no access to ARDS and countries who paid more for the oils countries were cheaper oils that determined the fate of countries that determined the fate of industries that determine the fate of technologies. But when it comes to something very basic, meaning like for example, internet, most people will argue with you today. That internet is a basic human right. Everybody should have access to that because if you deny internet. To folks, then they will not be able to grow and, you know, develop their nation or community or a company or family. I believe in order to access the internet, you need some kind of Silicon laptop, cell phone infrastructure, mobile infrastructure, some kind of. You know, wireless infrastructure, some kind of servers, some kind of cloud, all of that is built on a single technology today. Semi-conductors so if you believe that internet is a basic human, right? I would say semiconductor by definition is also a basic human, right? Because if you deny semiconductors and there's value, we believe everybody should have internet. Yet this nation or that nation where we will not let them have any semi-conductors. Well, if that is the case, then you're automatically saying that you're not let them have any internet. So this is a reason why we say semiconductor is the next oil, because we feel that it is a foundational technology that will run all the technology in the world. All right. The cloud infrastructure, mobility, schools, banks everything to do with communication, entertainment, everything. If it is that central to our economy is centered or life, I believe it is as central as oil has been in the last 100 years. That's the reason why we say it's crucial for human development or the next hundred years Over the last 50 years, we have seen lots of war breakout over oil. Do you think that this is inevitable with semiconductors as well? Or can this issue be avoided? If. Properly educated. No human go wants to go to war. This is my fundamental belief. Humans like to exist in peace. Most humans do, and they will like to find a way that we can all live together in a peaceful fashion. Now, typically when they feel threatened that Oh, My giving this technology, somebody else means I myself or my other citizens of nation or my children will not have access to the technology that creates anxiety. That creates leads to all kinds of apprehension. I think the best way of avoiding scenario like that, this way of widespread availability of these things. For example, nobody in his sane mind will tell you today that just because your children have internet, it means by children cannot have internet, right. That could be completely insane statement, but yet suppose that was the case. Suppose Nebraska having internet meant California cannot have internet because internet was made available only to six States in the United States. That let's say that was the case. Only six days could have it. We don't know which six, what the remaining 54 will not have internet. Okay. Now let's see how internet will play out. Now you asked me the question. Would people die because of internet? I will say yes. Because these six States are going to the heck. I don't care. I want to be one of the six. I don't want to be one of the 54. I think the way we avoided that is that in the early days of internet, we came up with the infrastructure so that we could grow it extremely efficiently around the world. I believe we need to do the same thing for semi-conductors for all foundational technologies is going to be like that. Like climate change is an example of that is one nation to nation. Three nation can do something. It actually not, I think semi-conductor should be such foundational technology that we should all share, grow together. Collaborate. Make sure. No nation feels threatened or denied, and they should not feel that if the nation, if their citizens are not given this technology, they will be left behind, or they will have problems in their hospitals or schools or basic facilities or growth of the nation. So I believe we owe it to us, our citizens of the world that we may just like we did with internet. I think we did it very responsibly there a lot to be a lot more needs to be done there, but I don't believe that I get up in the morning worrying about. God, there are some people in Africa who will not have any access to internet. Has Africa will forever be a dark continent. I don't believe that is the case. I think Africa has long way to go, but it will get better. I think somebody can, should be seen in the same way. So I'm curious to know what do you think we'll get. The semiconductors into these countries that really need them into the poorest countries, is it have to be some mental or incentive shift on the corporation side, maybe some sort of political action that needs to take place. How do you foresee us getting semiconductors to every single person in the world? I think the first of all, anything movement like that starts with awareness, the countries around the world have to be aware that just like internet. Just like clean air, clean water, and other human drives that they fought for. They will have to also fight for basic technologies. And if they don't fight for those basic technologies, their citizens will not have it. So the first is awareness. So every nation should have this awareness that this is the basic fundamental technology that I should have access to. Second, it starts with education. I think. All major universities around the world are now capable of teaching this technology. This technology is now quite fairly available. We are trying to make it even more available as we speak through set up of many open source foundations, education, trusts, and many other work that that's going on. But I think it's the first is awareness. Second is education Curt, I believe is when students going to masters level or PhD level. They should have a way to experience building of these chips. I think that is where I am focused on right now. How can a person sitting in car can imagine one day that I want to build a chip for my community and be able to do that at a cost, which is affordable, no cost low cost, and then deploy that chip to build some system for their community and solve a local problem. So that's the third step. And the fourth step is then many of the companies in these countries start using locally developers, sponsor and support local chip companies to use their chips, to solve their local problems. And I think that is happening. Already the reason for that is because this entire movement of, of the solution towards the edge, which we call edge computing is forcing people to bring computing and communication closer to the devices and machines that are connected to the real world. And that means thousands and thousands of use cases. That also means that we cannot sitting here in China or us figure out what somebody in Canada needs, which means local solutions will emerge. So I'm very hopeful that I think if you take these four or five steps from awareness to education, the ability of the students and researchers and startups to do chips, and then local development of solutions will lead to wider availability of Silicon Silicon solutions. And then it leads to what we call a democracy of Siliq. So you're the chairman of the Silicon Federation. Can you tell us what you guys do? Silicon Federation is an umbrella group and that was. Set up so that we can do exactly the kind of thing that we are talking about. We can help support open source movement, help support technology availability, to various different nations around the world. And especially we are more caring about the nations, which are poor nations nations, which have been left behind. And I think those are the nations, which probably need more help and support. So that is a, another area. The third area is education. We really believe that the first step is awareness, but the second step is education. The young people around the world do not care about this. They do not get involved in that. It's going to be a tough road ahead. I think somehow getting the young people excited about Silicon involved in semiconductors is a very big battle that we need to, this is a battle of the minds that we need to win and make it as exciting as software, make it as exciting as entertainment and any other field. It is exciting. I mean, I've spent my entire life. I believe there are a lot of exciting problems that are there for young people. We have just not done a good job of getting them excited and telling them what about it so that we can pull them into this field? Well, you're absolutely right about that. I can say that from my college classmates, all of us were obsessed with making tech startups. I'm going to be heard the word tech startup that was. Almost synonymous with software. We didn't even think of getting into the chip manufacturing space. We saw electrical engineering is kind of the boring department and computer sciences where you could make cool stuff and where you could get rid. Yes. And that's very unfortunate because no software runs without hardware. So you can do whatever you want to is software. You can actually, it has to run on some computing platform which runs on some kind of Silicon base you know as of today I wanted to go back to something you mentioned earlier, which is as part of the silicone Federation, and you're trying to educate not only the population, but also these corporations and trying to democratize Silicon. I'm wondering isn't the presumption that these corporations don't want to share their patents or their secret technologies because they want to, you know, protect their profits, protect their business. How do you realistically incentivize these companies to basically democratize and open source? Chip manufacturing and chip development actually amazingly open source helps even the large corporations, because if you think about that in open source, what is happening is a very large group of people are developing a piece of software or together. Or a piece of hardware together, but what it does, it brings down the cost of R and D. It brings down the cost of support and maintenance. It finds bugs faster. And as a result, the overall, you know, path of innovation, speed of innovation increases. So you will see a major corporation like IBM, Intel, Microsoft AMD, they support open source. They will not be supporting over source. If they felt that is counterproductive to their corporate goals, they have found out over time that the right mix of closed source and open source is the right way to go. So I think what we need to do is to use that force. To our benefit at least get basic foundational technology in the hands of many people around the world. Now what it means they will not be able to make a school stuff as at Intel or Microsoft or EMD or Xilinx for they'll be able to make some stuff. And that is okay with these large corporations, this large corporations not bothered about if you make. You know, some basic technology or some basic products, they're okay with that because they want to sell high value. So I think I actually, it works that we develop basic open source technologies, ordinary people, companies around the world will be able to use that to build chips system for themselves, but anything on top of that, which requires more complexity, advanced features or anything else. That requires significant more R and D that will be done by commercial corporations. And hence you will have to pay money to get access to that and has they can make money doing that while the basic technology is available to a larger population. So I think this right mix is there. That's why open source is supported. Well, literally all major corporations today. I think this is very interesting taking the open-source model and bring it to hardware because if we look at the analogy and software. I mean, you could talk about before open source and after open source, open source software created a Cambrian explosion for startups and innovation and all of these things. So right now, one of your hardware, we're going to hear chips. We think Intel AMD Nvidia. Do you think that open-source Silicon will create a new era of startups in hardware? Absolutely. I believe that is inevitable and I think. the major barrier that wasn't the basketball. Most people thought that it's so difficult to design a chip. It takes such a long training. We have proven all that to be false. We have taken high school students. We are taking university students with three months of training and a passion for Silicon. They have been able to design simple chips. And they have been able to take them out. Those tips are back there, put them on boards that are actually working today and, and thanks to companies like fabulous, Google others, have they supported these kinds of efforts, but more and more companies are not coming out. You will see more and more students, researchers, startups being able to do this very low cost chips. You're talking about. You could get a small chip done for less than$20,000,$10,000. This chip. Then you put it onto a board and then you can play with it, do something with it, whatever problem that you're trying to solve. This is not something that is outside the range of most people. So I think we have reached a point where this is doable and it's a matter of now educating people who are interested in startups in young people. So this is an option. And I believe if people get excited about it and they see their chips working and, you know, whatever robot or IOT system or some edge computing system, they are building, which is doing something cool and interesting, I believe it will lead to more and more people doing it. So I think the lowering of the barrier of doing it wider education. Open source movement. All of that will combine towards more and more people get excited. And I think you will see a lot more startups using this approach to build more interesting and innovative Silicon, bring them to systems solutions. Now, as I think that will happen more and more now, what do you think the next two decades of chip development will look like? So, first of all, I think that there will be an emergence of new materials. I think we will definitely. Start moving over from Silicon. I think we have pushed Silicon too, as much as we have. And yes, there is a few more generations maybe, but they'll become extremely expensive and probably not viable. But this has been said about Silicon for the last 30 years. And every time when you say somebody would come forward, okay. You tell me three nanometer is not doable. Let me show you how to do one nanometer. So this. I consider that as a challenge statement rather than a disability statement. So I think that will continue. But at the end of the day, I think there are other composite materials that can be used on. And I think data science has made a lot of progress in that direction. Secondly, as the Silicon moves out of a box out of a cloud into the real life, Gets embedded into human body gets embedded into animals. Get embedded into machines, get embedded into our furniture, into buildings, into our lights, into our life. That select that piece of computing communication doesn't have to be Silicon. It could be other materials, more friendly, easier to process, maybe less expensive to process. Then I believe this whole Revolution that is happening in the energy space that is going to impact our industry quite a bit as new materials emerged, which are a lot more efficient than the standard solar cell based on the PV. I think that is going to impact. And the third thing is that as I think we start worrying about the remaining 4 billion people, which so far we have not worried too much about. And we tried to raise there. You know value of, you know, quality of life and pull them out of power D in Africa and Southeast Asia and other things like that. I think that would put an extreme pressure on costs. And I think we'll be forced to reconsider that, you know, how can we build cell phones, laptops, all kinds of other machines and things like that. A very low cost. I think all of that will. Lead us to find alternate ways of finding computing done with different materials, with different manufacturing processes. And so I look forward to the next two decades, a very exciting time. We are lot new innovations will happen then, and we will explore beyond Silicon, but we'll also explore how we can power the world. Without having to burn fossil fuels and, and, you know, currently solar energy already is the cheapest form of energy available on planet earth today, as we speak, nobody would've thought that possible 30 years ago it is possible now, but this is the, we can still, that still is the beginning. We think we have a lot more to go. And the scaling of this technology can help us even get much better. So yeah, I think next two decades are going to be extremely exciting and none of the new innovations to come in this week. So what do you think needs to happen to get young people, young engineers and young entrepreneurs to be some excited about hardware? I think the first thing we must do is that I think Harvard industry has done poorly. When it comes to advertise itself, I think most human beings do not appreciate what hardware does she has done for the world. For example, if you were many people have done this comparison, it is little silly, but true. If you compare a car industry with SU with a chip industry, Today, the cars would cost probably$2 and probably would run, you know, 5,000 miles on a gallon. Something similar kind of numbers have been stated. I'm just trying to make a joke out of it. But my point really is that that is what we have done in the chip industry. Yet most human beings. If you talk to them all over the world, they will not be even able to tell you that that, which industry. Has done this kind of remarkable achievement in the last 30, 40 years you ask them, which industry do you think? I would be very surprised if less than 1% people, more than 1% people will actually say, Oh, I think it was the semiconductor industry. So I think 0.1 is that we have done a very poor job of VI bean hardware industry has done a very poor job of educating the ordinary citizens about. What remarkable things we have been able to do and how it improves the quality of life of everybody around the world. I think we haven't done again, a very poor job of that. Secondly, I think I believe universities have not done a great job of attracting students because what was happening, that exciting thing never happened in the university. It happened after you joined a company. You never taped out chips in the university. You never built interesting robots other than MIT, Carnegie, Mellon, UC Berkeley, USC, and a couple of other universities. I said USC, because I know you're from here. Yeah. I have to appreciate that USC USC, for sure. So, so my point really is that beyond these few universities, We don't have that cool factor that, Oh, you can go to electrical engineering and computer engineering department. Somebody's building some robots, somebody doing some other things based on the chips they have to add. I think, you know, and the reason for that was because you didn't have the capital and funds and things like that. So as a result, it was too expensive. So I think we are trying to bridge that. And I think the third very important thing is. That somehow, at least in few advanced countries, things which are considered complex, like chip design considered geeky became non fashionable. Some 30 years ago, 20 years old. This is geeky only Greeks to that as if you know that that is something that we have to as a society overcome that, that, like, for example, I think Elon Musk has done a great job of making rocketry, an interesting field. And nobody says that, you know, it takes a. A rocket scientist to do it well, normal ordinary people work at SpaceX and they are doing great and it is letting other people, there are actually some 60, 70 rocketry companies around the world now. And I think this is, I think that is great. And I think similar kind of stuff has to be done for chip design here. We just make it something cool. Exciting. Interesting. And I think that you know, how can you spend 40 years of your life doing it? I said that the reason for that is because there has never been a single day where I was not excited about doing what I do. Fantastic. But so I have not done a good job of translating that and a great job of telling others how exciting it is. So w one thing which I think is maybe quite telling of this is when I was in college. Most of the people who are majoring just kind of consider that like their work. But the computer science kids consider that their lives. I don't think there's any other major where people had, would have stickers in their laptops for things related to their major, where kids would literally in their free time go to hackathons would make their friends from other people within their community. Like software engineering at the university level is, is a tribe. It's a way of life. I don't think, I don't think electrical engineering or hardware, or really any other engineering major has managed to get that level of excitement or passion. Yeah. And I think one of the reasons for that is because the barrier to enter in a Sophia is very low. You have a, if you have a software, you have all the tools, you have a laptop, sorry, you have all the tools that you need. And so laptop and a passion is that what you need. But in case of a hardware, you need a laptop and a passion and a lab. And I think that is that many universities, either defunded those labs or didn't have them or reserved them for only PhD students or master's students. And certainly undergraduate students didn't have access to such last year. They can put up things. And secondly, I think open source movement didn't come soon enough. We, where we had tools software tools to design these chips available nausea. So I think increasingly I'll give you an example. There is a, a lab in Pakistan. It's a small university is called postman university of technology. So it's monitors, it's you, if you look at, in the, even in the map of Pakistan, it will not show up in probably top a hundred universities in Pakistan. It's that small, they set up a lab off. Five six, 10 veggies undergraduate student, two professors who had a passion to build a risk five processes. On one of my trips to Pakistan, I took the two inventors of risk five with me, Dr. and we were giving a talk. And in, in my my university entity university, these guys came and says, can you come with us? We want to show you something. They got into a car. And their university has maybe 30 minutes, 15 minutes away. We went to their lab. It was a small lab, but on the wall, they had all these risk five posters as if it was like a temple. Believe it or not, they taped out a risk five processes or months ago, no resources, undergrad students, but based on the passion and open source, they taped out the chip came back. It is the first microprocessor design in Pakistan. It's very simple. It's not very complicated, but doesn't matter. They designed it. They have it, it is working and they are, cannot stop talking about it. But again, they're undergraduate students. This is what I say. Now, this lab that we have in the Swan Institute of technology can be anywhere in the world, how much this is, why I'm. Nothing zero cost. Just your laptops. You sit there. They just found they didn't, the university refused to give them some space. So they went to the top level, the roof, they themselves put some stuff on top to cover it. And then that became their lab. It was like, no air conditioning, nothing there, but that's what they did. But my point is wow. With a laptop. Passion undergraduate student and two professors to guide them. They're designed the first microprocessor from Pakistan. Now, having them designed in FBG, there are several other universities in Pakistan. Pakistan is not known for. Semi-conductors just soy is not in Brazil like that is chili and neither is many other countries. If it is beginning to happen there, I'm very hopeful. It will start happening everywhere. That's. Yeah, Dr. Sherwani, I think you've made a fantastic argument for why more people should consider hardware at the high school level, university level. So if we have to switch topics a little bit, I wanted to touch on AI. I recently wrote a book called AI superpowers by Kai Fu Lee. And you had a lot of hot takes in that book by the way. One particular hot tech that he had, he suggested that in terms of AI innovation, China's far ahead of the United States. So in your experience having worked in China and having worked in the United States, would you say that's true. Is China sort of the leader or the follower when it comes to AI? Well, I think first of all, when you say AI, it's a very broad field, right? So, so there are many different aspects. There are many sub-segments and there are many fields inside the AI. Right? So to I think when you're writing a book and you're, you are talking to a broad audience, I think it's a reasonable comment to make. But where I sit, what I I look at is I say that what are the leadership in each segment? So I think where China has done well, is that when a segment has been defined and the path has been made clear, and all that was needed was to put in sheer hard work to take it to its logical conclusion. China has done very well, right. The areas where you have to open up new pathways and open up new innovation and things like that, where I would think that us has still some lead there. So it's a, I would not call it a, a one zero win for China. I think there are areas where. China is winning and there are areas we are us is winning. And I think it's actually clear why they are winning. It is it's kind of a cultural divide. The cultural divide is there are things when you just need out of sheer hard work with innovation and all that that's one. And then there is just here opening up new pathways and things like that. That's a different thing. So I think both countries along with others, Europe is doing very well. Japan has been actually the original innovator and the sustainer of AI they are doing well. And I think I see both countries not catching up in other areas. So I use your model there. Both countries and, and other hostile countries will do very well in the AI area. So I would not be as blanket as saying that China somehow has a lead in AI. I think there are some areas that China has the lead, and I think there are some crucial areas in which AI in the U S has the lead. Okay. If I had to give you something more concrete, maybe conversational AI, right? I think a lot of people want this. I think it has Universal usage potentially. Does China get their first or does American get their first rubbery? China. Full self driving. Probably U S why do you say that? Because of Tesla? No, no, no, no, no, no, no. Tesla is probably one of them. There are many, many folks in us. We're making very, very good progress today. And there's an explosion of investment. Universities are involved. We, we have a very broad network of folks working on it and, and us. So I would think that probably fully self-driving cars are neither necessary. Or not needed. I think, I think what is needed is what I partial. Self-driving. I mean, if, for example, if you just automate our highways, I think there'll be more than sufficient for the next 30 years. And, and we can drive the cars in the cities, as long as the highways can be automated. I think that will solve a lot of problems. So I think those are the kinds of things you will see earlier in us than you will see in China. Of things is always in China. For example, if you, I, I visited a conference about two years ago, almost by accident, where there was a kind of an exhibition or competition was going on. We are different companies are showcasing their autonomous vehicles from a golf cart to a ambulance, to shopping cart and lots of things like that. And they will competitions going on. And that was almost two, three years ago. I did not see anything like that in us at that point. But when I looked at each one of them, each single competitor, I did not see that level of advancement yet of it was there. So I think this is a bit versus depth question. So. So, yeah, so I, I, that's the reason I think that I have a feeling that autonomous driving is on us probably to get there earlier than other missions. So regarding ML, to what extent do you think that this is a hardware problem that is these things innovations you've described and democratizing Silicon? How will that benefit the ML space? Well, I think in terms of ML, We have to allocate resources, appropriately hardware resources. So what happens at the edge? What happens at the cloud and what happens to the bandwidth in the middle? There are three fundamental technologies that will play a role. So, so you have to have 5g so that you have enough bandwidth from the cloud to the edge. You have to have enough compute power so that you don't have to send a lot of data. And you should be able to do a lot of ML in the edge itself. And then you have to have a fairly gigantic cloud infrastructure, which has to be deployed at a reasonable cost around the world so that you can, the data that does come to cloud, we can infer from it and then we can send it back to the edge. So I think ML is a hardware and a software problem. I think, appropriate development of software. Necessarily, but hardware, not just in the compute, but communication infrastructure is necessary to make sure that the ML progresses you know, in future as rapidly as we wanted to. That requires lot of, I mean, if you look at edge, computing is already growing at a very interesting rate. It's a$250 billion business already today, but I think it will continue to grow. I've seen amazing demos of agricultural drones. Making decisions on the fly and, and lots of other things like that. And, and then I, I think that is very, very important that we have those kinds of capabilities and their ability to communicate with other drones. And then that drones, swarm ability to communicate with the cloud, I think, is very necessary. All that is different kind of chips or communication or computing chips would be necessary for ML to continue to grow. Okay. Dr. Sherwani You've suggested some pretty futuristic ideas and products you see, happening in the future of the moonshot projects of the unicorn ideas that are realistically possible in the next 20 years. What excites you the most? I think what excites sometimes means they're already there, but I think I would like to rephrase the question. Is that what I would like to see? Sure. I would like to see. Global education infrastructure for the world. I think I would like to see each human child have access to education and opportunity if we can not do that as planet shame on us, because I believe that fundamentally will take many generations forward. If we consider every child as our child. Every person as our brother or sister around the plan and treat them as we will treat our brothers and sisters that requires creating an education and opportunity network around the world. I believe today we have the resources to do that. And I think semiconductors and software, this is the two fundamental technologies that can make it because education is such the it and hardware and software and a willingness special. And vision to think of world as one. So I think to me, the most important thing that we can do for this world and which we should do. And I hope to see that we do is that we make education available and considered education as a fundamental right, of ever every human being, every child that is born on planet earth. And then once the education is there, then creating opportunities for them also globally. So I think that has to be the biggest mega project of humanity. For the next 20, 30 years, because that I believe will bring peace prosperity, but more importantly, I think we can live without having to worry about this and that this nation and that line. And we can all live in peace and harmony. And I think that will come from the creating education and opportunity for all. That's a really great answer. I was gonna say that's a great answer. When me and frauds were discussing this exact question, when we're prepping, we're thinking like, you know, robots and I robot or like robots that can run really fast and fight in Wars. But this is a much better answer. I mean, rewards and all those will will happen. But I think if as a, if we cannot read all, see right now, look at. Fundamental issues around the world. They all resolved around one human, treating another human as not equal whites, blacks, this nation, that nation, they were all of that. If you can just actually start thinking. And I always tell people that if I can start thinking that every child is my child, I'm willing to do a lot for my children. Why am I not willing to do for other people's children? Because they are other people's children. When are we going to cross this chasm? Everything. No, no, no, no, no. There's nobody children. They're all my children because they are, we are living on a shared planet earth those days, where we were living in a cave. And I only had to worry about my family or my driver long gone, but I unfortunately feel that we are still living in that scale. We still think there is a difference between a white and a black and this nation and that nation. And just, you happen to be born on six inches on this side of the wall, you become an Iranian. And if you want six inches on that side, you become a Pakistan. There could be a hospital, which on this side of the border of that sort of border and your mother could have been on either side of the border and it decided your life. It decided your future. It decided to opportunities decided your education. Very unfortunate. So that's the reason I believe that our goal as your entity has to somehow put this tribal past behind us and the quicker we do that, the better we will be better. Our future generations will be, unfortunately, most humans don't agree with me. They like to cling on to their tribal past and, and try reality, blinks many, many things, but. I dunno, it brings some kind of safety or feeling of belongingness to something, but that also creates divisions around the world and it causes all kinds of issues. And you see the disasters created by those divisions. Yeah. I, I completely agree with you. I think you I think your mission is very admirable, and I kind of imagined that too many people would disagree that it's one worth pursuing. So as software engineers, as people in tech our audience, who's listening to this. What can they do to help this vision of democratizing education opportunity for people in the world? First of all, I think people who are working in open source are already doing it by making it available to everybody all doing that. Secondly, I think all of us not just become students, but also become teachers because when you're a teacher, you're a better student. So I think there's so many opportunities now to set up classes where you can start teaching you know, there was a conference on open source software are done by open source software foundation of Pakistan, 125,000 people logged in. I gave a talk on risk five open-source technology, 12,500 people logged into my car. I've never done anything that impactful where I was talking to 12,500 people and trying to convince them that they should go back, set up a small lab, like Koosman Institute of technology and tape out a chip. Why not? Why would you not do that today? All your listeners have that ability, all your listeners. Have the ability to not just learn, but also teach. So I think that's one. Secondly, I think we should develop an empathy for people around the world. I think that requires learning about other people around the world and their condition, life condition and things like how can we help by the way, I must say that for younger generation, I found a major difference between you and us. I referred to an older generation. We were too concerned about my car, my home migrates, my job, my salary. I've seen already a very hopeful sign in next generation. Many of your viewers, my own daughter, and all of you guys, you care a lot about the world. You care a lot about planet. You care about lots of these causes, which we didn't care. I'm talking about me as a generation, not me as the individual. I think I already see hope. I think you guys are very caring, I think to you, it matters you guys on the front line on fighting a lot of the social ills that we have, which we fought, but not with as much passion as you guys do. So I'm very hopeful. I'm I have great trust in our new generation that you guys have already taken on you. I think if you dedicate yourself, Expanding your mission, reaching out to nations around the world, teaching growing people. And you can do that today. Sitting in your home, you don't have to travel, tends to pandemic. We have figured out how to reach out to the other side of the world. I think all that, all your viewers have that power in their hand today. Come teach, join, improve lives of people around the world. The most important it will do for you. It will improve your life. It will make you a better human being. And if nothing else, if only one human being can be made better, I think this world would be a great plus. Well, I think that's an exemplary way for us to wrap things up here. So the last question that we always ask is, in your opinion, what is the best piece of software built either in recent history or Oracle? Well, I think I, I, I made sure not, I should not answer this question mainly because. I'm not an expert in that area. I have very limited exposure to software. I'm a mislead, I'm Indian, a hardware person in my life, in, in my life that Sophia that I have interacted with, I would say that is a Linux because to me, that opened up a whole field. That was probably not open before that. So. I believe that has done wonders and then the entire software stack that was built on top of that and that what we see today, all the virtualization that happened in the cloud. I think all of that started from there. I believe that is a fun foundational piece. And then what Linux foundation has done actually on top of that, That whole Linux movement. I would say again, I have limited knowledge. I think that was one piece of software started and created into a movement, has done a humongous thing for humanity. I think we can not be, you know, thank them enough for what they've done. I think Linux actually is a proof of the mission that you're pursuing right now, right? Like no one would have thought that an operating system, maybe the most complicated piece of software in the world would be built for free. We'd be accessible by anyone could be maintained by people in Europe, in Asia, in America and South America, anywhere in the world. And yet they have, so if we can do an operating system, And supported by all for profit companies. Right? Right. That's another thing you ask Microsoft and IBM, do you hate Linux because they are cutting into your profits? No, they will tell you. We love Linux. That's amazing. Nobody would have thought that would be their answer was no, no, no. We will do everything to remove Linux from planet earth. That's not the case. They are participating in it. They're supporting it. They're helping it. Red hat became one of the largest acquisition. I mean, I believe 40 odd billion dollars, something, something amazing thing like that. So I believe humanity discovered something, then there's joy and profit in sharing. We just didn't take that. So I think I always knew there is joy and our mothers and fathers told us, Oh, there's joy of sharing. I changed that statement a little by saying there is joy. And profit. So that's my only modification to that statement. And I used to tell my mother that you always told me that, but let me tell you something. I discovered joy and profit in sharing, and I believe that it's true. And I think because joint collaboration models. But bringing more innovation, it just, it's just not possible for 300, 500, 700 people to sit inside the company innovate, which is a whole humanity can sitting all parts of the globe can do it. It's an incredible force that you will be utilized to solve all major human, you know, problems that we have in humanity today. I think it would be such a tragedy that we did not use the collective. You know, brains of all the human beings around the world. And I think open source is a very good vehicle. I think you've made a very convincing argument for that. I just want to add Dr. Shawnee alerts a lot. I mean, I'm really not that familiar with hardware. My dad just quick shout out to my dad who used to take me to circuit city and take me around the shop and show me every piece of hardware to explain it to me. So he himself is a trained electrical engineer and you try to like enforce upon me how important hardware is. And as a kid, and as like, even as a high schooler, I would never pay attention. But now I have to go back to the top. You have to go apologize. Yeah. Give your dad a call and say, I'm sorry. I don't think you need him an apology. I think what you need to do is to assert that to other human beings, that how can we learn more about hardware? Because I think when you learn more about hardware, you will also come up with new ideas for software because you can innovate. Why have you always organized CPS this way? Why do we organize GPU's this way? Can I. Design, this FPG in a different way. You never had that choice. You are stuck with the, with the FEG as we gave you. We wait today, you have that power. You can come join our open source FPJ foundation. We'll teach you how to design a Fiji of your own. Then you can write software for it. Now your combination of software and hardware might do 10 times better than whatever FPGA or GPU or CPU that you had before. So I think if we can get folks like you get excited in something like that. Well, I'm going to make my software more powerful by actually trying to fiddle around with my hardware. I think you will find new and amazing areas to innovate. That's our episode for this week. Thank you so much for listening. Make sure to subscribe to us and rate us on Apple podcasts. We would really appreciate the support. You can also follow me on Twitter at F Z from Cupertino and Busan. The ad next facade. See you guys next week.