Pod Insights The Brand Called You Silicon Dreams: India's Semiconductor Journey | Prof Ram M Vemuri, Prof & Head, Dept. of Electrical & Computer Engineering, Mahindra University About this episode In this episode of the Brand Called You, Prof Ram M Vemuri, a seasoned semiconductor industry expert and current professor at Mahindra University, shares his fascinating journey from IIT BHU to Intel Corporation. The discussion covers the evolution of the semiconductor industry over the past three decades, touching on technological advancements, manufacturing challenges, and the global supply chain. Dr. Vemuri also provides insights into India's emerging semiconductor industry, discussing the challenges faced and the role of educational institutions in preparing the workforce. About Prof Ram M Vemuri Dr Ram M Vemuri is the professor and head of the Department of Electrical and Computer Engineering at the Mahindra University in Hyderabad. After his Ph. D. Dr. Vemuri joined LSI Logic Corporation working on the development of the then industry-leading Electronic Design Automation (EDA) tools. He has worked with Intel Corporation for 25 years.
Transcript 0:08 Speaker 00 Welcome to another episode of the brand called Hugh, a podcast and podcast show that brings you leadership lessons, knowledge, experience and wisdom from thousands of successful individuals from around the world. I'm your host Ashutosh Garg and today I'm delighted to welcome a highly respected professional from the world of academia from Hyderabad, India, Dr. Ram Mohan Bhumri. Dr. Bhumri, welcome to the show.
0:32 Speaker 01 Thank you, Ashutosh.
0:34 Speaker 00 Dr. Bhumri is a professor and head of the Department of Electrical and Computer Engineering at the Mahindra University in Hyderabad and he has worked with Intel Corporation for 25 years. So my first question to you is can you share your journey from IIT BHU to the semiconductor industry and what inspired you to pursue this field?
0:57 Speaker 01 Certainly, Ashutosh, it would be a pleasure. So I always wanted to be an engineer even in my very early childhood. I used to tinker around with a lot of things and my parents always used to score me because I would open up things without knowing how to put them back together. But anyway, I did end up, okay, becoming an engineer. So I completed my BTEC in Electrical Engineering from IIT BHU and after completing it for a short while, I actually worked at Bharat Petroleum, which was only a few months. And very quickly, I realized that that was not the kind of work I was actually interested in. So I came back to IIT BHU to do an MTEC in Systems Engineering.
1:39 Speaker 01 And those days, Systems Engineering was the closest thing that I could get into that had anything to do with computers and I was interested in computers. So after completing my MTEC, I actually went for a PhD to the University of Cincinnati in Ohio, USA. Where I completed first my MS in the area of signal processing. And then my PhD in the area of VLSI engineering. So at Cincinnati, while I was working on my MS, I did a project on three dimensional noise cancellation. And while I was doing that project, as I was implementing it, I actually implemented it using
2:22 Speaker 01 a DSP or digital signal processor from TI, Texas Instruments, which was called the TMS thirty two or twenty. And while I was doing that, I my curiosity got aroused that how do you build these chips? So once I completed, of course, I didn't want to leave my MS in the middle and just pursue my curiosity. So I completed my MS and then I actually went and studied VLSI engineering, which is the same as a chip design.
2:53 Speaker 00 Right.
2:53 Speaker 01 So I did a PhD in from the same institutions, in University of Cincinnati. And this time it was in VLSI engineering, I learned how to design complex processors and how to build these complex processor, what was the art and science involved in coming up with something as complex as a DSP. And obviously from there, okay, there was a question of pursuing it as a profession. So I joined in nineteen ninety four, I joined LSI Logic Corporation. So LSI Logic, which doesn't exist anymore, it was bought out by Awaia, which was later also acquired Broadcom.
3:36 Speaker 01 So now LSI Logic became a part of Broadcom and we all have heard of Broadcom as the company. Few years of spending my time there, I actually moved for personal reasons from the Bay Area to Portland, Oregon.
3:52 Speaker 01 And Portland, Oregon is a very beautiful city and both my wife and I both fell in love with the city and we wanted to move there. And I tried to get into Intel and Intel was very happy. And we both clicked and so that's how I ended up in at working at Intel. And for the next twenty five years, I worked at Intel and rose through the ranks and I retired in twenty twenty one.
4:18 Speaker 00 Wow.
4:18 Speaker 01 From Intel as a senior manager. And then I did want to come back for a long time. It was a dream for me to come back to India and contribute to the country that actually gave me so much in my early years. So now I'm fulfilling that dream of mine.
4:34 Speaker 00 How amazing. Incredible journey. Thank you so much for sharing this incredible journey. So, uh, you know, you've been involved in the semiconductor, you know, area for a very, very long period of time. And I remember many, many years ago, we used to have an India manufacturing facility in Chandigarh, which then then didn't really go very far. I wanted to ask you my first question relating to the industry, which is how has the industry evolved over the last three decades? And what are some of the major technological advancements that have been very impactful?
5:11 Speaker 01 So over the last three decades, okay, it has grown by leaps and bounds. It has grown and changed by leaps and bounds. Okay, so when I first joined, uh, L S I logic, we used to measure the transistor sizes in terms of microns. And micron itself is very small, but when you look at the current industry, okay, micron is considered as magnitudes of orders of magnitude larger than what we see as device sizes today. So it has undergone a lot of evolution. I would say not changes per se, but it is more of an evolution. And
5:51 Speaker 01 In those days, when I was first getting into this field, there were no tools to design these chips, okay? Much of the work used to be done manually. And as you can imagine, when you're doing things manually, it is not going to be easy to do, design something very complex. And in fact, the first computers that I remember used to come with sixty-four kilobytes of memory, okay? Which is insignificant these days. Today, we talk about memories in the ranges of terabytes. So that it has gone through a lot of evolution. And most of this evolution is actually driven, okay,
6:35 Speaker 01 by the fact that the manufacturing process itself has continued to shrink the size of the devices. So there are two different factors that come into the picture. One is the hunger from the users to want more and more power in a smaller and smaller footprint.
6:55 Speaker 00 Correct.
6:55 Speaker 01 Right, so today, the first computers that I used when I went to the University of Cincinnati used to be huge behemoths. We used to have two computers. One was the VAX seven eighty-five, and the other one was VAX seven fifty. Both of them used to be in a separate room, okay? With the
7:14 Speaker 00 With air conditioning.
7:15 Speaker 01 Glass, air conditioning and the glass walls so that people could look through and see those behemoths. But today, what you and I carry in our pockets is orders of magnitude faster and more complex and can do a lot more than those computers put together at that time. So what is driving this? The driving forces, like I said, okay, there are two things. One is the users wanting to have more and more power in a smaller footprint and with a battery that can last for a much longer time. And on the other hand, the manufacturing process itself, the semiconductor manufacturing process, which is shrinking the device sizes
7:59 Speaker 01 so that a large design that used to fit maybe
8:06 Speaker 01 On a much larger die space, okay, could now be put into, okay, one tenth of that space. So when you're giving that kind of flexibility, people will use as much as they can get. So, yeah, so I have an anecdote about this, okay, in the very early days when I was still working at LSI Logic, we used to measure, normal for, for normal chips, we used to measure the, the number of chips or number of dies per wafer.
8:36 Speaker 00 Right.
8:37 Speaker 01 So a wafer is basically a round object, okay, with a semi, of silicon, on which you create the chip that you're going to
8:46 Speaker 00 Right.
8:46 Speaker 01 Put into a package and sell. So we used to measure it in terms of number of dies per wafer. Until, at that time, IBM came along with one of their very big designs. And manufacturing it was so difficult in those days, that we used to joke around that now we have to measure the yield, not in terms of dies per wafer, but wafers per die.
9:14 Speaker 00 Wow.
9:15 Speaker 01 So, yeah, so we used to have to, the, the big, the chips were so big, that we could not actually design them, we could not manufacture them on, uh, every single wafer. It used to be, take several attempts before we could actually get one working chip out of those. So this is how the industry has transformed. And another unique feature about the semiconductor industry is that in most areas, you see that the industry follows academia. But in this particular field, industry is leading academia in many aspects.
9:54 Speaker 00 Wow.
9:55 Speaker 01 There are some aspects where academia is also leading as far as theoretical aspects go, but because of the cost of starting a manufacturing facility is so high, okay, academic institutions cannot afford to have those. So a lot of this very Cutting edge technology is being developed at the industry and then the academia is coming in and finding out theoretical understanding of how these things work. So there is a leapfrogging effect where we keep leapfrogging over one another, where academia leapfrogs over the industry at some point and industry leapfrogs over the academic institutions at some point.
10:39 Speaker 01 And as I see it, it's not ending anytime soon. Okay, so the drive for more and more power with the artificial intelligence applications and learning machines coming into the picture, intelligent machines coming into the picture. Okay, things like chat G P T, all these things are continuing to drive. Okay, the demand for more and more compute power with less and less.
11:08 Speaker 00 Interessting, thank you. What a great response. The other perspective I wanted to ask from you sir was about India. You know.
11:20 Speaker 00 There's a lot of discussion and a lot of action going on to be able to manufacture and support our own semiconductor industry. What are some of the primary challenges we face in India? And what can the government do to support this?
11:37 Speaker 01 Oh, very good question. So the semiconductor manufacturing requires a very efficient global supply chain today. Okay, and with designs often being done in one geographical region, while the manufacturing might be done in a completely different geographical region. And everything that is required in between from to go from a concept all the way to the manufacturing being spread out all over the world. Okay, so this supply chain is an extremely important aspect of the semiconductor industry because there are many points in the supply chain.
12:17 Speaker 01 And one thing that COVID nineteen, very clearly, the COVID nineteen pandemic, which very clearly highlighted was the problems associated with this supply chain. So during COVID nineteen, it was, in those days I was still in the U S and it became extremely difficult to even buy a car.
12:41 Speaker 01 And not because the auto manufacturers were not making them, but because the chips that were required. Every car has over hundreds of hundreds of chips in it. So the number of chips that were required and the supply the chain, okay, breaking down because of COVID nineteen pandemic. Uh created this uh uh uh famine basically for things like automobiles and refrigerators, washing machines. All these, okay, which run on semiconductor chips, okay, became very difficult to obtain in the market during those days. And that is one of the factors. Then there are many other factors and some of them are geopolitical tensions.
13:26 Speaker 01 So for example, US and China, right, or India and China. So the the relationships between these countries, as long as it is a good relationship, things can prosper. But once the relationships start uh seeing some stress, okay, we also start seeing that the semiconductor industry is suffering. Because parts, many parts of this are actually coming from China. TSMC, Taiwan Semiconductor Manufacturing Corporation, which is the largest semiconductor manufacturing in the world, okay. Is somewhat controlled, okay, because of the relationships between Tai- Taiwan and China, okay, it's more controlled by China.
14:05 Speaker 01 And getting chips from them is also, uh our manufacturing shifts and giving them, okay, our technology is also something that uh many countries are avoiding these days. And then, okay, there is the the complexity. Okay, so when I, when I say complexity, the complexity of the manufacturing process. These are now, okay, we are talking about nanometer scale devices and nanometer scale wires. So putting all these things to a chip is a very complex process. And there are only a handful, not even a handful actually, of manufacturers that make these equipment.
14:46 Speaker 01 And this equipment itself, which is required for the semiconductor chip manufacturing, is also a very, very advanced technology. So in order to set up a manufacturing facility, the cost for a cutting-edge leading technology runs into about five to eight billion dollars just for the setup. And the amount of time from concept, from the time that you laid on the foundation to the time that it actually gives out its facility, runs into at least two to three years. And in order to get it to a volume manufacturing, it takes even more amount of time. So it will be close to five years from the time that the fab is built
15:29 Speaker 01 to the time that you can actually expect volume manufacturing out of that facility.
15:37 Speaker 01 And when it comes to India, your question was specifically also to do with India. So, yeah, all of these problems that I mentioned, they are also applicable to India. But in addition, okay, we have a few more problems that some other countries, more advanced countries, are probably not facing. So one thing is the lack of the manpower, the human power, resources that are required, trained resources that are required to manufacture, to fill the manufacturing positions that are required. Right? So we, we are very good, okay, when this comes to design.
16:22 Speaker 01 So we do have, in India, we do have a lot of design houses. Because design does not require this kind of investment. Design is basically, as long as you have some computers and some of the software, which is not as expensive as billions of dollars, okay, you can actually design. We're very innovative people. Indians are very innovative people. We come up with a lot of ideas. What we are lacking is, okay, the training to manufacture. So design, we are very good and we have all the resources. So many of the Institutions, academic institutions, some of the leading institutions like I A S C or some of the I A Ts. And private institutions like Mahindra University.
17:05 Speaker 01 We are doing, okay, what we call semiconductor workforce development. So we are working on it. We have in Mahindra University, for example, we have a B tech program. In V L S I design and technology. And we also have an M S program, M tech program, which is called V L S I. Design and embedded systems. So we are focusing on some of these areas and there are. A few handful of other private institutions that are also looking into filling this void, okay, that is being created. By India's desire to be self-sufficient and to have indigenous manufacturing within the country.
17:44 Speaker 00 Very interesting. But you know, in the last few years, I've been reading a lot in the papers that there are three or four major projects. To manufacture semiconductor chips in our country. With a lack of talent or non, not, not, you know, non availability of talent. How will some of these projects be able to sustain themselves?
18:09 Speaker 01 So. We live in a global village, right? We are not isolated anymore.
18:16 Speaker 00 Correct.
18:16 Speaker 01 So if you look at. The Tata group, for example.
18:21 Speaker 00 Yeah.
18:22 Speaker 01 Okay, they recently hired Dr. Ranbir Thakur. Okay, he used to work at Intel. So, in fact, he was in the same group. He was at a higher position than I was and. Later on, okay, he moved on and Tata brought him over to India. So along with him, he brings in a lot of talent. So people like him who have done this work abroad maybe, okay, that is something, okay, that we can tap. People of, especially of Indian origin, they're, many of them might still. Okay, have their hearts in India while physically they might be in a different country. So they are looking for ways to contribute to the Indian, the India's development.
19:07 Speaker 01 And secondly. Like I said, we are not, we're living in a global village.
19:12 Speaker 00 Hmm.
19:13 Speaker 01 And what I mean by this is, there are people we can tap. They don't have to be of Indian origin. They don't have to be in, living in India in order to, for us to tap into their knowledge, tap into uh their expertise and learn from what people have done. Because this is a very complex industry, so a lot of mistakes happen along the way. So if I can avoid, like they said, that uh wise people learn from the mistakes of others.
19:42 Speaker 00 Okay.
19:42 Speaker 01 Right, so we are in a position where we can learn from the mistakes of others.
19:47 Speaker 00 Hmm.
19:47 Speaker 01 So people who started this, they must have made many mistakes along the way and learn from those mistakes now. They can teach us so that we can avoid those pitfalls and our sailing can be smoother than theirs.
20:00 Speaker 00 Correct. And you mentioned a few minutes back about Mahindra University developing programs and working on educating students or preparing students for the semiconductor industry. Do you see a bigger role for educational institutions in our country to be able to prepare? Because with such a large demand as you just told us, with every, even every car needs hundreds of chips. Uh, do you see educational institutions giving a big thrust for training engineers in this particular area?
20:32 Speaker 01 It, it's a cart and a horse situation.
20:35 Speaker 00 Okay.
20:36 Speaker 01 Yeah, so students are generally interested in uh studying in areas that will, that have promising careers.
20:46 Speaker 00 Absolutely, yes.
20:47 Speaker 01 And in order to build those careers, okay, we also need to be able to create those jobs.
20:53 Speaker 00 Hmm.
20:53 Speaker 01 So it's a catch twenty-two situation because without the expertise from the uh educational institutions, we cannot start a manufacturing facility, for example. And until we start a manufacturing facility, okay, the jobs are not there. They are future jobs. So we are.
21:11 Speaker 01 What we do is we take some minor steps like I mentioned earlier. We India is very good at conceiving designs and bringing them all the way up to the point of manufacturing. We don't have manufacturing yet. So our focus is to introduce the students. So to train the students so that they are not left jobless after they graduate. We need to make sure that they have a career immediately after graduation. And also, so, so from that perspective, we teach them the design aspects. And then from the second perspective, which is the number of future jobs that we are expecting, which is going to explode in the manufacturing industry for semiconductors.
21:55 Speaker 01 We introduce them to that. So those of the students that are actually interested can either go and start getting some of those jobs that are already becoming available and build up their career based on that. And as that grows, okay, then more and more jobs will be created and we'll see that students who were this interested only in the design aspects might start showing more interest towards the manufacturing also.
22:21 Speaker 00 Very interesting.
22:22 Speaker 01 So unless there is a student interest, we will not be, we can run the program, but we will not be able to fill the seats. So it's a delicate balance where we have to make sure that the programs that we are creating here are looking at the future, but they're not forgetting about the current.
22:40 Speaker 00 Fascinating. So sir, time for one more question and I'm going to ask you this about a company which is making waves all over the world, which is Nvidia.
22:50 Speaker 01 Oh, yes.
22:51 Speaker 00 What is Nvidia doing, which is so different, that an established organization like Intel is, I mean, I was reading somewhere that they've Nvidia is talking to acquire Intel. What is it that they're doing and what is it that they have done so well that we will not be able to replicate in a hurry in our country?
23:15 Speaker 01 It's somewhat of a, I need to be politically correct also here. So I need to, yeah, I need to.
23:22 Speaker 00 And, and if you, if you think I shouldn't, you shouldn't, don't want to answer.
23:26 Speaker 01 No, no, no, no, absolutely. It's a good question. So the reason I'm saying politically correct is in some ways it is luck.
23:33 Speaker 00 Hmm.
23:34 Speaker 01 So I, I don't know how well, uh, how far back, okay, you were interested in this arena.
23:42 Speaker 00 Hmm.
23:42 Speaker 01 Okay, but Nvidia used to make graphics chips.
23:45 Speaker 00 Yes.
23:46 Speaker 01 Okay, it was only known for its graphics chips. And even when I was a graduate student, when I, this was uh, three, four decades back, okay, we had artificial intelligence. So it is like a perfect storm coming together. So in those days, okay, compute power was not there.
24:06 Speaker 00 Hmm.
24:07 Speaker 01 And AI did not take off because it required immense compute power.
24:11 Speaker 00 Correct.
24:12 Speaker 01 And then Nvidia, okay, it was starting to somewhat struggle in the graphics market.
24:18 Speaker 00 Hmm.
24:18 Speaker 01 And then, okay, the advantage they had was in a graphics processor, you have a very good multiplication unit.
24:26 Speaker 00 Hmm.
24:27 Speaker 01 So a multiplication unit is very fast multiplier within the CPU or a GPU in this case. And then, okay, because they had this uh, very good, very efficient multiplication unit within their GPU.
24:41 Speaker 00 Hmm.
24:42 Speaker 01 And there were more and more requirements from artificial intelligence and machine learning. The two came together at the right time.
24:50 Speaker 00 Hmm.
24:51 Speaker 01 Okay, and at the same time, a third factor, okay, that was developing independently was the FPGA market.
24:58 Speaker 00 Hmm.
24:58 Speaker 01 So FPGAs were becoming more and more efficient so that accelerators, which are a big part of uh, any fast processing, especially things like AI and machine learning.
25:09 Speaker 00 Hmm.
25:10 Speaker 01 Okay, they could be put on to these FPGAs. So all of these came together and they gelled.
25:18 Speaker 01 And of course, Nvidia was extremely smart in the fact that they recognized what power they had in their GPUs.
25:27 Speaker 00 Mm hmm.
25:27 Speaker 01 And the fact that, okay, this is what is required for machine learning. And they did a terrific job of continuing to support that area.
25:36 Speaker 00 Hmm.
25:37 Speaker 01 And now they're the market leaders, okay, as far as processors for anything to do with machine learning.
25:45 Speaker 00 Correct.
25:46 Speaker 01 And another thing, myself, okay, being from Intel, I know that there was a burden that Intel always carried.
25:54 Speaker 00 Mm hmm.
25:54 Speaker 01 And which actually restricted its growth into new markets.
25:58 Speaker 00 Hmm.
25:58 Speaker 01 And that was backward compatibility.
26:01 Speaker 00 Hmm.
26:01 Speaker 01 So Intel had a very large ecosystem of software that was already developed for the x86 architecture.
26:10 Speaker 00 Hmm.
26:10 Speaker 01 So in order to continue those software to run on newer processors, we had to make sure that any new processor, any new architecture that was being developed at Intel was still backward compatible with the x86 architecture so that, okay, all the existing software suddenly, okay, wouldn't stop running on the newer processors because that would be a market killer.
26:36 Speaker 00 Yeah.
26:36 Speaker 01 So there are similar situation existed for AMD also, AMD and Intel, both of them use the x86 architecture, so they were both burdened by this.
26:47 Speaker 00 Hmm.
26:48 Speaker 01 And whereas Nvidia came into the picture much later, and they did not have a burden that they needed to carry, legacy burden.
26:56 Speaker 00 Amazing, amazing. Thank you for such a great response and for educating me and a lot of our viewers and listeners. But on that note, sir, thank you so much for speaking to me about your own amazing journey in this very exciting field. Thank you for speaking to me at such length about the semiconductor industry, some of the challenges, some of the opportunities that are faced. I just love the way you handled the whole issue of the geopolitics, which I know plays a significant role in this very key component in everything that is electrical or electronic in our, in our lives. Thank you also for speaking to me, sir, about Nvidia.
27:38 Speaker 00 And Intel and AMD. Thank you again and all the very best.
27:43 Speaker 01 It was a pleasure, Ashutosh. And I would like to, although I am close to retirement now, I would like to see this industry continue to prosper and bring us fascinating new things and new horizons that we didn't even imagine possible at this time.
28:01 Speaker 00 I'm sure it will happen. Thank you, sir.
28:05 Speaker 00 Thank you for listening to The Brand Called You, videocast and podcast. A platform that brings you knowledge, experience and wisdom of hundreds of successful individuals from around the world. Do visit our website, www.tbcy.in to watch and listen to the stories of many more individuals. You can also follow us on YouTube, Facebook, Instagram and Twitter. Just search for The Brand Called You.
Previous Changing Landscape of Investment Banking and Corporate Law | Rosalyn Breedy, Managing Director, Breedy Henderson Ltd Next Evolution of EdTech | Alexander Zheltov, CEO & Co-Founder, Educate Online Inc Page updated November 15, 2026 · transcript and summary are machine-generated