No Innovation Without a Market: Ahmad Bahai on TI’s R&D Model

0:00Building a new fab is a five-year-plus kind of initiative, I mean, getting it, and then multi-billion dollar investment. So if you want to address the next pick, you have to start building in the trap.

0:13Peter High: Welcome to Technovation. I'm Peter High, President of the Strategy. My guest today is Dr. Ahmad Bahai. Ahmad is the Senior Vice President and Chief Technology Officer of Texas Instruments, a global semiconductor company that generated nearly$18 billion in revenue in 2025. TI designs and manufactures analog and embedded processing chips that support applications ranging from industrial equipment and automobiles to consumer electronics and AI infrastructure. Ahmad guides TI's breakthrough innovation, corporate research, and Kilby Labs connecting emerging technologies with opportunities across the company's businesses.

0:46Peter High: His career includes research leadership at Bell Labs, founding semiconductor company Algorex, and serving as chief technology officer of national semiconductor. He's also a professor of the practice at MIT and an IEEE fellow. Today, we'll explore how TI turns research into commercially important products, the role of analog technology in powering AI, and the advances needed to bring intelligence to the edge. We'll also discuss how materials, packaging, and manufacturing are expanding the possibilities for semiconductor innovation. Ahmad, welcome to Technovation. It's great to speak with you today.

1:19Thank you.

1:20Peter High: My pleasure. Thank you for inviting. And now for a word from our partner, Blitzy. Want to shift software five times faster? Meet Blitzy, the autonomous software development platform built for enterprise-scale initiatives. Your code base has outgrown any one team. Tens of millions of lines, decades of technical debt, more than any other AI tool can hold in its context window. Blitzy starts by understanding. Thousands of specialized agents map your architecture, dependencies, and business logic across every language into a knowledge graph. That graph also services hidden CVEs and zero days before an alert ever reaches your desk.

1:56Peter High: Then you describe what you need in plain English and approve the agent action plan Blitzy builds. Grounded in the knowledge graph, Blitzy executes entire engineering epochs over weeks of continuous compute, and every line is tested end-to-end before it reaches your team. Modernize legacy systems, close critical vulnerabilities, ship new features five times faster. Learn more at blitzy.com slash technovation. That's B-L-I-T-Z-Y.com slash technovation. And now on to the interview. It's a great pleasure. I'm looking forward to the conversation, certainly. Ahmad, I thought we might begin with TI and the business that you're in, certainly a storied business across many decades now.

2:40Peter High: Bring us up to speed on the business as it stands today in your own words, if you would. Yeah, thank you again for inviting. It's a pleasure to have this conversation with you, and your introduction covers a lot of bases on TI, so I'm going to elaborate on some of these details. As you know, Texas Instruments has been around for quite a while, more than 95 years, and we are the largest analog mixed signal company in the world. Pretty much every device that has an on and off switch, there is a good likelihood that you have a lot of TI chips inside of it. And we are covering a wide range of products, technologies, and, of course, market targets and customer bases.

3:19Maybe one of the most diverse companies in all of these aspects. As you know, analog and mixed signal is a critical part of any electronic system, ranging from consumers to medical to industrial, automotive, space, aerospace, and defense. because at the end of the day, the world is an analog world. You have to sense things. You have to amplify signals. You have to deliver it to a machine that can do the signal processing and turn it to digital and make it an AI-enabled or any other applications you can think of. So from that viewpoint, we have the diversity of technologies. We have the diversity of products, diversity of customers and markets.

4:01But also another angle which is very unique about TI is that it covers all the different aspects of the chip manufacturing and production, meaning from design to manufacturing to packaging to even distribution. And that makes it very unique in the sense that you have an end-to-end presence in this market. Not only is it a resilient business, but also is very critical business for a wide range of applications, as I mentioned, in different markets. And having these manufacturing capabilities in a way that you can have a very reliable supply chain, because you are pretty much in control of your destiny, pretty much, by having that kind of capabilities.

4:48Of course, that takes a lot of investment, as you pointed out, and you have seen in many news broadcasts and many other resources that we spend more than almost$60 billion on expanding our manufacturing. because we are a big believer that the demand for chips is going to grow. Some of the interesting numbers is that people had forecast about by 2030, the chip market is going to be more than a trillion dollars. Guess what's happened this year? We are going to cross that boundary already by far. So all the demands that we see out there and having more electronics, more intelligence, more connectivity around us, number of things that you carry in your pocket, in your hand, in your body, and in your car, and everything around.

5:40The infrastructure that we have in industrial, in AI, in defense, in the space, they all depend on electronics, and high-performance electronics, and reliable supply chain. And they're all kind of topics that TI spend a lot of attention to make sure that we can offer the technology advantage with a very reliable supply chain. So that's kind of a high-level view before getting to any specific topic on this kind of whole value chain.

6:13Peter High: Yeah, phenomenal overview. Thank you so much for that, Ahmad. As the Chief Technology Officer, you're responsible for innovation, corporate research, the Kilby Labs. I'd be interested in learning a little bit more about how you describe your purview across those responsibilities and perhaps others as well? Yeah, actually, semiconductor industry as general, I think, is one of the most, if not the most innovative technology in the human history, because I don't think any other technology or innovation managed to make it million times or more than million times more efficient at the lower cost or at the same cost.

6:49That is not, I'm not aware of any other innovation or technologies with that kind of success story. So we thrive on innovation and manufacturing and reliability and capability of offering this massive number of chips every year, billions of chips. So innovation is critical for us to stay ahead in terms of the performance, in terms of reliability, and in terms of new technologies for new applications. And this is a kind of ongoing race. It's not like you cannot just take pride on your past. You need to constantly innovate to make sure that you offer the most competitive and most compelling products out there.

7:33And when you have the diversity of technology from design to manufacturing, innovation is going to be a very multifaceted initiative. and you need to take into account innovation at the circuit design, innovation at the manufacturing, innovation at the device, new devices, which goes back to physics and how to bring the most challenging technologies and materials to the level of manufacturability, innovation on packaging. And all of these really covers a wide range of topics. And we will try to address that in TI in a very creative way, rather than separating our innovation engine from the rest of the company.

8:15We mix it together because innovation has to happen at every corner of the company. That's why our research lab is tightly coupled with almost every segment of the company, even the manufacturing and sales and tools and software and hardware and everything that relates to the entire business of chip production from design to pretty much distribution.

8:40Peter High: Hmm, fascinating. Can you describe, so you're referencing the extent to which innovation is mixed across the entire company. Talk about the team you lead and how much of it is directly tied to you versus dotted line connected to you as a result of some of that sort of federation across the organization, perhaps for a lack of a better way of framing it. Yeah, that's a very good way to put it. Actually, Kilby Lab, which is really a core research lab in TI, my office, in your office, is in charge of looking at high-risk technologies that need to be de-risked before getting it to the level of business units and production lines and other groups in the company.

9:24So we start with de-risking, and de-risking has many different challenges and different formats. For example, we can work with a partner outside. We can work with universities. We can work with many other resources. Once we have more confidence in terms of de-risking process of a new technology, we start working with the rest of the company very closely to make sure that that technology has a home. And there are groups that are heavily involved in developing these technologies. Because I'm coming from a Bell Labs pedigree. You know, many years, a couple of decades ago, Bell Labs was the place to do research.

10:01And one thing that we learned there is that in research, you cannot just pass the baton and just say, okay, goodbye. You have to team up with the people who are going to use these innovative ideas and make sure there's a one team mindset behind this research work. Many organizations, some of our competitions have this big research institute which work in isolation and having a big challenge to transfer their ideas to the rest of the company because that integration, that collaboration and teamwork doesn't start early enough to create that sense of, okay, we are part of the same game. And that's why we see the de-risking process and adoption needs to be tied together in a very careful manner to make sure that there is a seamless process in place.

10:52Peter High: Fascinating. So this is really, in essence, the way in which you ensure that new ideas can effectively be scaled, not just be a series of experiments that don't really see in the light of the day or move the needle on an already consequential and significant organization. Absolutely, because no matter how great the idea is, no matter how disruptive the innovation is, if there is not enough understanding of the implications for a product line or people who are going to take it to the market, it's going to sit on the shelf until it expires. You really need to make sure that there is, in the technical term, impudence match between the people who innovate and people who are going to take it to the market.

11:33And they all feel that they are part of the same kind of activity and the same kind of like engagement at different stages, of course, with different levels of commitments. But one team working from the early de-risking all the way to production with different emphasis on the resources or skill set that you need at every stage.

11:56Peter High: Semiconductor development requires long time horizons. I don't need to tell you. while customer needs and end markets can change rapidly. And so I wonder, as you think about the balance between the necessity for patience with the commercial, the near-term commercial opportunities that continue to present themselves, how do you think about getting that balance right? Okay, there are multiple ways to look into that. So there are some building blocks that are needed no matter what technology you're working on. When you're building an electric car or a data center or a drone or a robot, there are building blocks that are needed.

12:34You need to manage the power. You need to amplify the signal. You need to take an analog signal and convert it to digital. So these are blocks that we call it analog block, meaning that they are so fundamental that it's kind of like agnostic to some extent. I mean, not fully, but to some extent, meaning that no matter what you do, you need to manage the power. No matter what you do, you need to handle some of these signals and condition them. So, of course, each market has different spec and requirements, but the fundamentals are pretty much very similar. So in that case, there is a common denominator of technology that goes across.

13:12So we make sure that that portfolio of the generic catalog products not only is fully addressed by TI, but we have a very competitive and compelling solution for that one. That's why we see that some of these parts are always in demand, regardless of the ups and downs of different markets, because they are so critical for every electronic system. Then it comes to parts that are really pushing the envelope of technology. And for that, nobody can predict the future. If you know someone, please let me know. But in the meantime, what we can see that we can see the vectors and which direction vectors are pointing to.

13:55For example, we all knew that electronics at some point requires to address higher voltages because of the electric vehicle needs higher voltage or you need to have better performance on the battery life because of the robotics and some of these systems. Or you can see that these data centers are growing and they have some specific set of requirements when it comes to power, when it comes to timing and clock, when it comes to interfaces. So you don't know exactly when things happen, but you can see the vectors. And as you pointed out, technology development in semiconductor is a very complicated process.

14:32You need a lot of physicists and chemists and material scientists and manufacturing people put their heads together to come up with tiny devices that you can hardly see. So that is where we see that you need to intercept the technology and market by some understandings of the vectors. And you don't know exactly when it happens or which market is going to be the biggest or the most promising. But you can see the trends and you act on the trends. And then you adjust your timing accordingly. So you're not going to put all your money in one basket because something happens next year, but you can have some flexibility.

15:10So it's a very careful trade-off analysis of risk, reward, market opportunities, and engagement with customers at different level. Because in some cases, customers are talking about their immediate need. I need something tomorrow. Of course, you need to have it in your inventory. And sometimes they talk about the trends at CTO level that, okay, we see five years from now. Here is how our business is going to look like. So you have to translate those trends to opportunities for us and start building up the core technologies. You give an example. About 10, 11 years ago, maybe a little bit longer, we realized that high voltage is going to be a big market for us.

15:57At that point, we thought that opportunities are going to be mainly for smart grid or things like that. So we developed some technologies at transistor level for high voltage applications. With understanding that there are markets that we think are going to grow. That was a long journey. Halfway through that journey, we noticed that a new market came up, which is going to use this technology. That was electric vehicles, because they need high voltage. And then three-quarter way in that journey, we noticed that data centers are the biggest market. And then AI became. And so it's hard to predict any of those 15 years, 10 years ago, but you can see the trend.

16:35And then you adjust or adapt to different opportunities and markets. That's how I see it.

16:41Peter High: Yeah, fascinating. I want to talk to you a little bit about that, the compounding effect of each of these waves of innovation, obviously culminating of late with the remarkable innovation happening with AI and its relationship to the remarkable increasing demand for semiconductors and obviously the advantages that are parlayed to an organization like TI. Talk a bit about the AI revolution and its impact on your company, if you would. Of course, we could see that. I mean, AI has been in progress since 1950s. I mean, do you know that the first AI lab was in MIT in late 1950s by Minsky and his colleagues?

17:26I mean, so obviously we saw that trend happening and we always had our eyes on it. But obviously, over the last several years, especially in 2017 and all that, when generative AI started ramping up, and we see that this huge growth of AI, we looked at it in multiple ways. I mean, AI can help us to do a better job in our chip manufacturing, and we can do a better job offering chips for AI. So it's a very circular relationship for us, meaning that we help AI to become more powerful and AI help us to kind of make things more optimized and more efficient. A lot of our chips are being used in an AI infrastructure.

18:15As I mentioned, one of the big challenges of AI is power, how to deliver so much power in a rack in the data center. So we are the number one power management company in terms of the semiconductor content. And we are strongly addressing that market, working with the key players there. There is an interface, the data centers process data much faster, but also communicate data at much higher data rates. So we are working on some of these interfaces. As you run a much higher data, you need a much better clocking. So we are working on timing references. So there are many aspects of AI that you're addressing, and we see that it's growing very fast, and we're very excited about that.

18:58But also, we are taking advantage of AI to improve our design process and manufacturing process more efficiently, because one of the secrets, one of the, I mean, it's a kind of contradictory statement. One of the well-known secrets of the semiconductor industry is that scale is everything. If you have a scale in semiconductor, you can manage the cost and you can dominate the market. And on top of that, you're going to be sitting on an amazing ocean of data. Our manufacturing facilities creates tens of terabits of data every day in terms of the whole manufacturing process. So we are sitting on this massive data.

19:42We want to take advantage of it and we want to make sure that our scale can leverage AI to improve our competitiveness even more.

19:54Peter High: I want to go back to the point you've raised at the beginning of your response about power. Its availability is a constraint on the expansion of AI data centers, certainly. And I'd love to get your thoughts on where you see the largest opportunities to improve efficiency between the electrical grid and the processor itself. Yeah, actually, it's fascinating because power, everybody thought that power is already a solved problem. I mean, people were talking about DSPs and algorithms, but then all of a sudden power becomes a bottleneck in some of these things. And when we talk about bottleneck, there are many angles to that.

20:32One is power generation. And you can see that there is a huge amount of innovation is happening. People are talking about a wide range of power generating technologies. You know, these small nuclear reactors, I mean, SMRs, modular reactors. We see that people are talking. There is a lot of investment on even fusion technology, which was tried many times. And now people are investing a lot more because of the opportunities there. There are, of course, many other creative ways to generate power. That's one aspect. The other one is distribution of power. And third one is that how to get the power to the processor, because processors are now a lot more power hungry because of their capabilities.

21:17We are talking about multiple thousands of watts in a single processor, and you have multiple of them. You can imagine you can have equivalent of multiple big heaters on a single electronic board. And you need to deliver that power in a very efficient way because if you don't deliver it, not only it's going to be a loss of power or a waste of power, but also a big heat penalty that really creates a big challenge for this AI infrastructure or these big data centers. And from that viewpoint, we work with them very closely to make sure that our power management and power delivery to the processor is optimized down to the last 1%.

22:08To make sure that not only we generate least heat, but also we deliver the power in a way that the processor expects to receive it. It's a very tricky business in terms of bringing a distribution line all the way to a 400 volt or 800 volt in a rack and then deliver it at sub one volt to a processor. And during these translations, not wasting much power. That is pushing the power to the limit. And that was amazing surprise again that, you know, demand on the power integrity and power efficiency in these devices is just becoming a big engineering science challenge. And we are welcoming that very much because it's exactly in the core components that we have been working on for more than a decade.

23:00Peter High: Yeah, interesting. And continuing with a related topic, what technical advances are necessary to make sophisticated edge intelligence sufficiently inexpensive and power efficient to become ubiquitous? Yeah, I mean, on this concept of cloud intelligence versus edge intelligence, we can go on and on. I have opinions. We can spend another couple of hours on that topic. But I think in the future, we see that its intelligence permeates different layers of the devices. So it's not just the cloud. Of course, cloud is the main for the training. But also you see that there is a lot of intelligence on the devices.

23:36could be your laptop, could be your phone, could be even your watch and your devices. So it's a federated learning and federated intelligence. And each of these are pretty much very much in need of a very high efficiency power supply. Of course, a lot of these edge devices are running off of a battery. So you're limited by how much power you have. So you want to make it very efficient to extend the battery life, but also you get maximum out of your processing capabilities. So for many decades, we had a very strong position, the leading position in our management for computers, laptops, and so on.

24:23We see that we can take that and translate it to the opportunity for devices. I mean, these days, laptops have a GPU on it and they have these processors that are doing AI. So HAI is also as critical, but the level of voltage, the type of device, and of course the cost is very different from a rack that is orders of magnitude, more expensive and more demanding for power. So it really scales the topic in a very, it makes it technically challenging, but the kind of boundaries or the boundary condition for your design is very different than you're talking about a ring on your finger, which has some intelligence versus a cloud versus the laptop and so on.

25:08And they're all pushing the edge of the semiconductor technology to the next level, both in power and also in other aspects of it.

25:17Peter High: Yeah, that's very interesting. Thank you for sharing those perspectives. As vehicles and industrial equipment become more software-defined, customers have to make increasingly significant software investments around their embedded processors. And I wonder, how is TI evolving its software development tools and even partner ecosystem alongside the hardware? So, I mean, in many cases, we are the users of the software. These are software for designing a chip or software tools that you need to run a factory and things like that. So we are working with these infrastructure software companies quite a bit.

25:54But as far as the tradeoff between software and hardware, it's been always there. What we think is happening is that the hardware needs to be reliable, regardless if it's your electric vehicle or your laptop or your phone, enough that you are upgradable to the next level of the software. I mean, there is an over-the-air upgrade of software, even sometimes in your cars, in your phones. Of course, so often you get an upgrade on your OS and software that's on your devices. But the key is that hardware, and when I say hardware, it includes connectivity, it includes power management, it includes the entire interface.

26:34and all the details of the hardware needs to be reliable enough to accommodate these software upgrades. We think that, you know, again, back to my earlier comment that some of the products or most of our products in TI really are so fundamental that regardless of the version of software or version of device that you have, you need these parts anyway. You need to manage the power. You need to convert analog to digital. You need to interface different chips and make them handshake and talk to each other. You need to have a clock and timing. So all of these are so fundamental that regardless of each version of the software and which version of device you have, you're going to end up using a lot of those.

27:21I mean, interestingly, one of our business leaders was saying that when you open up these electronic devices, there is so many tiny parts that you don't have space to put TI logo on it. But they're all TI devices that are sitting on these devices around a big chip. And that's why we see that these basic devices being an amplifier or data converter or, you know, DC-DC switcher, they're all critical and agnostic to a lot of these end markets. Of course, as I said, with different specs and requirements.

28:03Peter High: Yeah. Yeah. Progress in semiconductors isn't driven solely by shrinking transistor dimensions now. And I wonder how much of the next generation of innovation will come from process technology, advanced materials, packaging and system level integration. What are your thoughts? Yeah, I mean, I gave a couple of keynotes over, I have given it over the last several years about that very topic that, you know, in the past, I mean, past meaning up to like about 10 years ago. So the main vector for innovation in semiconductor or the dominant vector for innovation in semiconductor was scaling. And at this point, we see that scaling is becoming one of the vectors, not the only vector.

28:44There are many other vectors that you need to push forward to address all the needs we are talking about. Because when you're, as I said, delivering power, you're not using the latest, the finest geometry of the transistors. In fact, you want to avoid that in many cases because you're dealing with higher voltages. So you need to innovate not only on the scaling, which is going to be getting to the point of almost like below two nanometer is going to be a different ballgame in terms of architecture going to CFIT and other nanosheets and other technologies. But also you need to push the new materials for applications that don't demand that kind of scaling.

29:27I mean, a good example of this is like silicon photonics, which is really a new way of communicating between processors, between racks, and between data centers at high speed. And that doesn't demand a very fine geometry. Now, the beauty of that would be you need to come up with new materials, new device structure, new ways of thinking. And these are additional vectors that are really exciting. So the innovation in semiconductor is getting highly diversified in the sense that you need to push new materials and new devices for new applications that really are not tied to the scaling only. And on top of that, you have packaging and all the other aspects of it that requires more innovative way of thinking how to attach fiber to the module or to the device.

30:18And all of these are bringing more and more opportunities for us to make a difference. Another beauty of these technologies is that since they are not facing the scaling, your manufacturing infrastructure is going to last longer before you get an update or upgrade. And these upgrades are very expensive, billions of dollars. And that's why we see that in our business, we see that our manufacturing would be well positioned for many years to come to address some of these vectors of innovation.

30:49Peter High: Very interesting. Speaking of materials, you mentioned material innovation. How do you determine when silicon remains the best answer and when other materials or packaging architectures create meaningful advantage of one sort or another? Of course, we push silicon to the limit. silicon is an amazing material and it's going to be there for years to come and the foreseeable future because at the end of the day there is more than 50 years of innovation behind how to manufacture devices based on silicon but also we see that there are some complementary devices that can enhance the performance of silicon because what we have learned over many decades is that how to do manufacturing at scale extremely efficiently.

31:33I mean, it's amazing how we can do manufacturing at this scale and making sure that the yield and throughput is so high. So we wanted to take advantage of these many decades of learning, which was built on silicon and CMOS technologies, and expand it to other materials. For example, some of these high-voltage materials or some of these silicon photonics-related materials or things like that. So we will have many new materials. I mean, some of them are more integrated with silicon, like silicon germanium for many applications, including RF and others. Some of them are new elements in the periodic table, like gallium nitride, and people have been using gallium arsenide, so they are bringing it closer to silicon, for example, using this GAN on silicon substrates.

32:23So the idea is that how to leverage silicon learning, but also explore new materials. So I don't think we are looking at replacing silicon in all of this. We are complementing silicon with things that are new materials that can do some of the specific jobs better.

32:41Peter High: You mentioned at the outside of our conversation, Ahmad, that TI is undertaking a long-term investment of more than$60 billion across, I think, it's seven U.S. fabs. Talk a bit about that investment and the advantages you hope to parlay from it. It's going to be more than seven fabs. I mean, the idea is that, so here's the challenge in semiconductor business. I mean, the semiconductor business has a growing envelope over the last many decades, but there is a cyclic trend on top of this growing exponential, meaning that every four years or so, you see some peaks and some troughs. So that is given because of these oversupplies and markets and all that stuff.

33:25So if you can think of its exponential envelope with some cyclic wave on top of this exponential curve, that is the historic trend in semiconductor. On the other hand, building a new fab is a five-year-plus kind of initiative. I mean, getting it and then multi-billion dollar investment. So if you want to address the next peak, you have to start building in the trough, which is kind of tough business, right? When the market is down, you have to start spending billions of dollars. But since we have confidence and we had this confidence for many years that this market is going to grow and we think it's going to grow quite aggressively because of all these new things that you mentioned from data centers to robotics to wearables to everything else.

34:17and a lot of our products are going to be so fundamental for all of these markets. So we have this confidence that we need to have more capacity for manufacturing. So as I said, one of the key elements of the growth in semiconductor is scale. Because when you have a scale, you can reduce the cost, you can offer the most competitive solutions, and it's going to be harder to compete with somebody who has a scale in this market. So again, we had all these elements in our head and we put it on the very well-articulated plan. But then you have to have the audacity of start investing when the market is down or when the market is not there to be ready for the next bid.

35:03So we took that because of the confidence we had in the trend and the type of diversity we have in our product and end markets. And it's paying off. We see that now our new factories are getting filled relatively fast and new manufacturing capabilities, which are state of the art in analog and mixed signal, all of which are in U.S., are getting more and more critical for our longer roadmap. And we are very excited about it.

35:31Peter High: I mentioned in the introduction, Ahmad, your past experience at Bell Labs, one of the seminal innovative organizations in American business history. I wonder, what do you draw from that experience that you continue to leverage today? What made that a special place and what are the ongoing lessons that you continue to leverage? Yeah, I mean, Bell Labs experience was a very critical part of my journey because as a fresh PhD graduate, you get to rub shoulder with the giants in the field and learn a lot from what's going on in this most innovative, like you mentioned, the seminal kind of one of a kind type of research institute, which we miss it dearly, actually.

36:22But we all understand that that model was not sustainable because Bell Labs was based on a monopolized market with almost unlimited budget. That doesn't exist today. So obviously, that model served us, not only U.S. technology, but worldwide technology immensely. I mean, transistors and lasers and everything, fundamental technology goes back to Bell Labs research. But what was clear when I was in Bell Labs is that in many cases, research was done just out of curiosity, not tied to an end market, which can be justified when you have no budget pressure or no timing pressure. But in this world that we have an extremely competitive, even at the size of TI, you have to be extremely competitive to make sure that you have the presence in the market.

37:15You don't have the luxury of, okay, let's do research for curiosity. And if it happens to have a market, that's great. If not, you can have a great publication or a great prototype. So what I learned from that is how to tie curiosity and intellectual capabilities closer to what's happening in the trends of the market. Because as we talked before, most innovative ideas takes time. It takes many years to take it from a concept to something that is tangible and ready for adoption. And none of us have this crystal ball that five years, 10 years from now, which market is going to be dominant and who is going to use your technology.

38:03But if you carefully analyze the trends and translate it to these new innovations, being on the material or device or circuits, that is, I think, the most interesting part of this innovation in this world. Because, as I said, in a time and environment that you have no limitation, you keep inventing until one of them ends up hitting the jackpot. But today, you have to be a lot more understand and intent on which areas you're addressing. And the other thing that I really found very interesting, I mean, during my career, I always had a foot in the academia. When I was in CTO of National Semi, I was in Berkeley, then I was at Stanford, then now I'm at MIT as a part-time.

39:02And that gives me a chance to explore what's happening in the environment that is less pressured by the market and more driven by curiosity and how to tie this to reality of the competitive world. So bridging that gap also is a fun adventure.

39:23Peter High: And do you have optimism that that union between industry and academia can bear fruit in some new or unusual ways? I mean, not to say that it hasn't in the past. Obviously, there have long been unions between the two, but your most recent experiences now at MIT, for example, or does it have you optimistic that it might take the place a little bit of Bell Labs, of having other institutions, great institutions that have this long-term perspective without necessarily the pressure of commercialization in the near term? I hope so, but I would say there is a lot of challenge there because, as I said, even in universities, even top leading universities in the world right now, like MIT and others, are kind of like working hard to make sure that there is enough research fund to support different initiatives.

40:16So even though universities or academia is like less pressured by the reality of commercial market, but funding and funding resources are critical for some of these key researchers because what was unique about Bell Labs was people didn't worry about funding. That's one thing. And second thing was that there was a critical mass of know-how in every field. You pick a field, I mean, days there, or new materials for semiconductor manufacturing. It was like tens of top-notch people in that department. It's becoming more and more difficult to have that critical mass in even a university environment because there are individuals who are highly credible in the field and some of the leading figures, but you need a critical mass to build that wave.

41:12So I would say, to answer your question, I think definitely there is a lot of potential, but it requires a lot more engagement from both sides to build that critical mass. In other words, a lot of this research work requires the patience of many years and many days and nights of hard work. And unless there is a good understanding on both sides, there might be a cycle of excitement and disappointment if there is no consistency. So I would say we need to have a virtual Bell Labs these days by combination of university strength and industrial strength.

41:57Peter High: Very interesting. Ahmad, you've lived through numerous technological innovation waves across the past few decades. And I wonder, what analogies do you or lessons do you draw from past waves that you believe apply to the latest one? I would say, as we discussed earlier, nobody has the crystal ball to tell what is the most compelling market in 10 years. But one thing is amazing. When technology innovation and market disruption coincide, then you see the biggest bang for the buck. If you look at what happened in smartphone revolution, it was the convergence of the market in early 2000 and technologies that were in progress for many decades.

42:46and just hit the very unique target of, okay, can I have so much processing power with such a beautiful display in my pocket? So then AI, I mean, again, I mean, as we talked, I mean, AI has been in progress for many decades, but all of a sudden, a convergence of new algorithms in late 2017 or 18, new algorithms and generative AI, new processor technologies, and new peripherals, power management, and so on, all of a sudden came together to enable a new revolution in AI. So I would say some of these trends are going to continue. As you pointed out, the federated AI is going to be one of the waves of the future.

43:39We are going to see every device getting more and more intelligent, and every device requires more processing, more power efficiency and more battery life and so on. So I would say some of these trends can permeate into how to design the next device, how to come up with a new material, how to come up with a new software and architecture. And the next big wave would be a convergence of multiple disciplines rather than just one discipline drives everything, driving everything, yeah?

44:12Peter High: Yeah, very interesting. I wonder, before I let you go, Ahmad, what do you read or listen to in order to stay current or to help you think about what might be next? You mentioned there are no crystal balls, and therefore the people you're reading or those who are influential to you don't have them either. But I wonder, you know, for others who might be interested in replicating a little bit of your intellectual diet, what does it consist of? yeah i mean the diet needs to be adjusted according to your digestion system of course as you know but i mean i would say i think one of the things that i really enjoy is expose myself to innovation in almost every discipline even beyond semiconductor i mean i spend a lot of time understanding what are the innovations in biology and biochemistry and other stuff because you never know that some of these opportunities may intersect with what you're doing and make a big revolution.

45:06So you need two types of exposure to information. One is in your field, you need to, of course, stay on top of what's going on. That's part of your daily competitiveness drill. But also you need to expose yourself to wireless ideas that may not directly intersect with your job tomorrow, but can trigger opportunities and innovative concepts that you never even expected. So I would say you need a depth and breadth.

45:36Peter High: A wise counsel, certainly, Ahmad. Well, thank you so much for a great conversation throughout this session, Ahmad. It's been a pleasure to learn a bit more about your career, the areas of focus for you and your team, your view of the future. I greatly appreciate the insights you've shared. Thank you very much. It was a pleasure and I enjoyed the conversation as well. Thanks for listening to Technovation. If you enjoyed this conversation, subscribe and leave a rating. It helps more leaders find our podcast. Thank you and I look forward to seeing you next time.