3D InCites Podcast
3D InCites Podcast
From IMAPS DPC 2025: How Artificial Intelligence is Transforming Industries and Society
What happens when AI diagnoses patients better than doctors? Where does artificial intelligence truly stand on the hype cycle? Is all this computational power actually benefiting society? These critical questions frame our fascinating discussion recorded live at IMAPS Device Packaging Conference in Phoenix, Arizona.
Join our expert panel featuring Hemanth Jagannathan (IBM Research), Mark Kuemerle (Marvell), and Kimon Michaels (PDF Solutions) as they tackle AI's most pressing challenges and opportunities. Their collective expertise reveals surprising insights about how AI is transforming industries while raising important considerations about its implementation.
The conversation explores AI's evolution from specialized technical applications in semiconductors to today's consumer-facing generative tools. Our experts draw fascinating parallels between AI and previous technological breakthroughs like laser technology, suggesting we've only scratched the surface of potential applications. They provide compelling examples from healthcare where AI systems demonstrate superior diagnostic capabilities by processing complex datasets beyond human capacity.
While acknowledging concerns around data accuracy, power consumption, and appropriate boundaries, the panel remains optimistic about AI's future. They emphasize that today's implementations represent merely the beginning of a transformative technology whose full impact remains largely unanticipated. Yet they also agree on applications where human judgment should remain primary – including, amusingly, matchmaking.
Dive into this thought-provoking conversation to understand why organizations must either leverage AI effectively or risk being outpaced by competitors who do. Subscribe to 3D IncItes Podcast for more cutting-edge discussions on technologies shaping our future.
Interconnects for Tomorrow’s Applications
Disclaimer: This post contains affiliate links. If you make a purchase, I may receive a commission at no extra cost to you.
Become a sustaining member!
Like what you hear? Follow us on LinkedIn and Twitter
Interested in reaching a qualified audience of microelectronics industry decision-makers? Invest in host-read advertisements, and promote your company in upcoming episodes. Contact Françoise von Trapp to learn more.
Interested in becoming a sponsor of the 3D InCites Podcast? Check out our 2024 Media Kit. Learn more about the 3D InCites Community and how you can become more involved.
This episode of the 3D Insights podcast is sponsored by IMAPS, the premier global association for microelectronics advanced packaging enthusiasts. A membership in IMAPS helps your company grow its advanced packaging workforce through professional education and networking, advances your brand and supports building relationships. Imaps helps you learn, connect and collaborate. Learn more at IM imapsorg. Hi there, I'm Francoise von Trapp, and this is the 3D Insights Podcast. Hi everyone, this week we are recording live from the IMAPS Device Packaging Conference in Phoenix, arizona. Now AI continues to be a hot topic of discussion at industry events like this one, and this week several panels were dedicated to discussing topics such as are we going to have an AI winter? Or, as part of the Global Business Council, how are we going to scale AI from data centers to consumer applications? So here to talk about the key takeaways were some of the panelists. From each of the panels I have Kamef Jaganathan of IBM Research, mark Pemberley of Marvell and Kim and Michaels of PDF Solutions. Welcome to the podcast, guys.
Mark Kuemerle:Great to be here. Thank you very much.
Francoise von Trapp:So we've got a lot of questions to go through, but before we get started, can you each tell me a little bit about your company and your role there, Kim and?
Kimon Michaels:Sure, Kimon Michaels. I'm one of the co-founders and EVP of Products and Solutions at PDF Solutions. Pdf Solutions is the big data analytics platform for the semiconductor and electronic supply chain.
Hemanth Jagganathan:Yes, I'm Hemanth Jagannathan. I'm from IBM Research. Our overall charter at IBM Research is to be part of the future of compute. We work on all three elements of compute, which is bits, neurons and qubits. So I'm currently in charge of the Chippler Advanced Packaging Research Group at IBM.
Mark Kuemerle:Okay, and Mark. Hi, I'm Mark Kumerle. I'm VP of Technology for Marvell. I'm responsible for architecture for all of our custom products and also defining a roadmap to make those products possible.
Francoise von Trapp:Excellent. So let's dive in on the topic of AI, which is driving this industry. I want to start with asking you where would you say AI is on the Gartner hype cycle at this point?
Kimon Michaels:In our industry. I think it's still on the uptick in the beginning. Manufacturing semiconductors, semiconductors in general, is not the fastest moving industry and I think we're still getting traction. We're not in a trough yet, which invariably will come.
Francoise von Trapp:So for using AI in our industry, we're still in the up cycle.
Kimon Michaels:I think we're still in the up cycle, but how about as a driver of our industry? You know people are starting to get concerned about the draft. If you look at the panel last night, worry about is it the AI winter? I was at an investor conference where people were very interested in the NVIDIA presentation more about the future, not about them continuing to have excellent results. So there's certainly a concern. It remains to be seen.
Hemanth Jagganathan:I think we're still very early in the era of AI, because, while traditional AI and the solutions out there are driving the industry today, it's still in its infancy. We need to talk about an ecosystem of design, architecture, having a variety of options for an AI need, for example. So the industry is still very much driven to a few specific hardware solutions and then everything is running on that. So I would say there's still a lot of innovation and diversity to come.
Francoise von Trapp:I think there's a lot of improvement that's needed in what's out there. Do you feel like we kind of rushed things a little bit? Generative AI is the big thing right now and people are using it for making pretty pictures, writing emails, querying Google. These are all applications that aren't going to hurt anybody if something goes wrong, right, Other than the fact that you get not accurate information. But what do you think the best application is going to be for AI when we get to the point of really using it?
Mark Kuemerle:I have a controversial opinion that I think the consumer application and continuing to kind of scale up generative AI really has a lot of potential for the world. So when we think about, I mean there's so many places in our industry where we can use AI. You know, and people have been using reinforcement AI for a long time and in fact most of the place and route algorithms that we have for putting a chip together are all based on essentially what is early AI, finding local minimums and global minimums.
Mark Kuemerle:So to me it's not a new thing in our industry, even though everybody is like wow, ai is a brand new thing.
Francoise von Trapp:Okay, so there's a difference between in our industry and in the general public, right, but in the general public.
Mark Kuemerle:I think it's new to the consumer and gosh what an opportunity it has to affect everybody's daily life, more so than kind of using it to build more efficient widgets, which is something that we've been using software for for a long time. So I'm kind of excited about what it could mean to transform everybody's day-to-day tasks that they do.
Francoise von Trapp:Anybody else?
Kimon Michaels:Yeah, and I think, much like the invention of the laser, people didn't contemplate the many ways it could be used. I think the use of AI and its real impact is still unknown. It's going to be orders of magnitude higher than it is today, both in our industry, consumers, overall.
Mark Kuemerle:It's a fantastic analogy, I think another way of.
Hemanth Jagganathan:Also and I agree with what both panelists said I would say also that if you look at AI as a form of compute, just like previous forms of compute, what was available to a few is made available to the masses, and then it takes its own evolution in terms of adoption and its usability and how it's used. So AI is going through a very similar cycle. Like we in our industry, semiconductors has used AI for many years, you know, and it's, I think, something we rely on day in, day out, and its use in general public, in more enterprise solutions, is getting to be higher and higher. So the cycle of how new forms of compute get ingested by society as a whole.
Francoise von Trapp:Okay, so what industries do you think we're likely to see AI be adopted in most?
Hemanth Jagganathan:I think this question came up in the panel yesterday and the unanimous answer was there is no single industry which can be singled out. When you are talking about any form of computation which uses more efficient use of data to provide higher levels of insight and know-how, I think the impact can be across the board. So in the world of AI, for people who are on the fence, they either get to use AI or get used by AI. So you have to choose sides.
Francoise von Trapp:But you know, I think about this a lot and I am a skeptic sort of you know. At the same time, it's really made Google searching much more. You get much more information, but I do worry about the accuracy of that information and I did hear at one point that chat GPT has gotten dumber since it started because it's scraping Reddit, where you're not getting accurate information. And one of the speakers this morning talked about the forecasters predicting that AI will be smarter than intelligent beings somewhere between 2026 and 2030. But since AI requires training with accurate data, what are the chances of this actually happening, at least in the use that it's being used now by the general public?
Mark Kuemerle:I think you bring up a really interesting point, because models are getting larger and larger. One of our challenges, at least in the hardware business, is to make hardware that can fit these rapidly growing models. As the models are trained on, more and more data concerned that there's going to be a dilution that's actually caused by a feedback loop where the AI is being used to generate a lot of content, which is then being used to train the AI. So a lot of folks have a big concern and probably a relevant concern, that that could happen. So definitely understand to build these just massive models. It'll become more useful. We are going to have to think as a society how we're going to kind of vet the content that's out there and how these models are going to use it. Right now they just go after everything.
Francoise von Trapp:Right, right. Well, I would imagine in certain industrial applications such as medical and semiconductor, there are ways to ensure that we're actually using accurate data.
Kimon Michaels:Yeah, I think the one advantage in the technical fields is it's more data than language driven Right, so you can vet that your base data set is correct, I think much more easily than an LLM model doing general data, for example. Still there are challenges of the interpretation of the data and, of course, like many industries or technologies when they're new, I don't think we've stepped into the legal repercussions first.
Francoise von Trapp:Oh, that's a whole podcast all by itself.
Kimon Michaels:For example, you know, abs brakes, completely accepted everywhere, now released in the US several years after the rest of the world or Europe, because of concern of the legal aspect. If you can brake faster than the car behind you, who's at fault? And I think that'll be a natural kind of tap the brakes on the acceleration of AI in our industry, of getting to some of the concerns you have.
Mark Kuemerle:One thing that we shouldn't forget, though, is there's a lot of different use cases for AI. Right, as I mentioned in our field field, in hardware design, we've been using AI for decades. Right, I learned it too many years ago when I was in college. We even were leveraging early AI. I remember back in 2000, buying the textbook Multiple View Geometry, which was an introduction to computer vision back in 2000, where matrix multiplication was used to figure out what's going on in the outside world, leveraging cameras. So there's a lot of places where we don't need to be as concerned. Right, reinforcement learning is essentially solving a puzzle or playing a game that has a defined outcome. The winner of a system like that is always very clear. That's how models were trained to beat everybody in the world in a game of chess.
Mark Kuemerle:So there are lots of places where maybe we don't need to be concerned, and having more capability is going to just help solve some of these complex problems.
Hemanth Jagganathan:I think, from the impact of AI. There are many examples in our industry, as Mark mentioned, but also there are pretty phenomenal examples in the medical field, which ultimately, I think, are getting the core of is AI making humanity better.
Hemanth Jagganathan:Right humanity better, right, right, and I think there's a lot of promise where there are many medical conditions, where there's sparse data or very unclear outcomes of whether to go down a particular line of procedures for a patient, and AI is able to not only ingest all of the information which is out there, which comes at a staggering volume for no human to ingest, but then also, looking at the prognosis and the charts of a particular patient, to give a likelihood of what would be if one were to go there.
Francoise von Trapp:Right right, they can determine and advance what the success rate would be for, for instance, for a surgery or for a treatment plan.
Hemanth Jagganathan:It could help tailor a more custom plan and also predict the success rate if one were to go down that path.
Kimon Michaels:It's a great point. I read a scientific study recently that said doctors using AI have a much higher correct diagnostic rate than doctors alone. Right right, but AI alone was even higher, which is the contrapositive of your concern. Worrying about the accuracy of AI is one side of it. Not trusting it on data-driven areas is the other.
Francoise von Trapp:So the doctors didn't trust it and they were wrong.
Kimon Michaels:Apparently, in some cases they probably overrode it, where, if you did the AI straight up, it had a higher percentage of being correct.
Hemanth Jagganathan:But it comes back to the ABS question as well, in terms of litigation and how do you know if you're getting the right set of data. So, for the foreseeable future, there will always be a human override because of some of these concerns. And still it reaches the point of ABS, where there are a set number of rules and repercussions of trusting it.
Mark Kuemerle:But I think, Kimon, you've got a really great example with the medical, and this is why I think it's important for us to again kind of think about different use cases for AI in different buckets in our minds and how much we worry about them. Would you necessarily be worried about a medical diagnosis driven by an AI that was using all of your vitals and using a database and using a reinforcement learning algorithm that is very likely to make statistically correct conclusions? In that case, it's just the benefit to us is going to be so significantly outweighing the threat. What's the?
Kimon Michaels:real value of AI to your point. It's not doing what humans can do more efficiently. It's taking really high dimensionality of data and complex sets to get to the answer humans could not get to if you put 1,000 men in a room for 1,000 days.
Hemanth Jagganathan:And I think also, if you look at over the past few decades, the amount of data which an average human consumes has been growing at a very rapid pace. So that also plays into that right. As you get more and more data, you get overwhelmed with what is there and then you make gut reaction choices. So if you're wearing your Apple Watch or if you're having a health fitness tracker, it is in some sense processing all of the data and giving you a score You've walked enough for the day or you've had your Right. So taking that to the next level as well, I think AI will really provide that value as well. I agree.
Francoise von Trapp:So in your talk, kim and you were talking about generic AI and saying that it's not going to solve alone. It's not going to solve the problems of the semiconductor industry. So can you explain what you meant by that?
Kimon Michaels:Yeah, I think if you look at the advances in ML algorithms and AI in general, it's phenomenal. I mentioned in the talk we run a course at Carnegie Mellon and what undergrad grad students can do these days, even with data in our industry, is incredible. The challenge, though in semiconductor the data is physical in nature. There's physical relationships between it. It's not Gaussian distributed. It varies over time. It has a strong temporal component Two different sensors which may have the same name, you expect to have different values. So understanding this industry perspective of the data the temporal nature really drives to how you use AI. With the data, you want to model its continuously trait, you want to monitor its accuracy, not only over the parameters you built for the model, but every other parameter, because the next excursion may not be something encapsulated. So it brings the challenges of our industry too.
Francoise von Trapp:Okay. So one of the questions from the audience at last night's panel really stuck with me. He said that we've all seen the charts about how much power was consumed by AI in 2024. And the question was and we've talked about some of this now, I think but what has AI given back to society for all the power it has?
Hemanth Jagganathan:Well, I think there are efficiencies which are coming into the picture, for example, the enterprise space. By infusing AI into a lot of transactional processing, you're able to stop fraud before it actually happens. The usual credit card fraud was caught after a few transactions go through and then you stop that particular card. For example, by infusing AI into enterprise applications, you're able to stop it before it happens. So it saves a huge amount of loss to the industry by infusing AI there.
Hemanth Jagganathan:We talked about the medical field, where one is able to get a better view of the prognosis and also the proposed treatment plan. So there's a lot of benefit there. I think, in general, ai is also really helping in bringing the world closer together. But there's so much of data out there, also with natural language processing, which, if harnessed correctly, will be able to help connect different languages together much easier. It is something which we have to still be very careful about, because there's a very interesting correlation between the human brain and, for example, the ability to learn languages and the propensity for Alzheimer's, for example, if you are not having the plasticity of your brain maintained. So sometimes there are these very interesting questions by using AI, are you not making your brain as plastic Right, are we dumbing down society because they don't have to use their own critical thinking skills.
Hemanth Jagganathan:So I would say AI is given a lot, but we have to always ingest it carefully and keep monitoring it, because there's a short-term and long-term benefit and a reaction. So we'll have to just see how that goes and be responsible as a society.
Francoise von Trapp:So I have an opinion about this. Do you think that generative AI for consumer applications could be considered a frivolous use of AI that will negatively impact society due to the drain on the grid?
Mark Kuemerle:Well, I guess I started out talking about how consumer applications are one of my favorites, I know right, so I have to jump in and defend myself to that question.
Mark Kuemerle:There's a couple of key points I want to make here. One is that I do really feel that consumer applications are going to be really important for kind of empowering people to be more productive and empowering people to understand data that might be beyond whatever means they have or whatever experience they have under their belt. I also think there's a great opportunity for us, as people who are developing artificial intelligence or machine learning hardware, to do everything we can to improve the efficiency of these systems, and I spend the vast majority of my time trying to figure out how to do that. And I believe that by helping to make them as power efficient as possible, we can kind of reduce that load so that a greater number of people can use it more freely. And the other thing I maybe want to share is that, while it may seem like a power user in my mind, at least, my personal opinion it's a lot better than spending that power on cryptocurrency mining. I'm a big fan of it from a benefit to society versus energy consumption.
Kimon Michaels:I would agree with you there To that point there's nothing wrong investing in entertainment and arts.
Hemanth Jagganathan:That's across society.
Kimon Michaels:In the beginning. New technologies, especially when they're exciting, they're already always overused until people figure out really the correct application, et cetera. You know, you look at music in the late 70s, early 80s way too much synthesizer, Right, it was new. Then People went over the top.
Francoise von Trapp:To that point. Are there applications where you think that AI shouldn't be used? There is a dark side and nobody talks about that, but we have to be really careful, right.
Hemanth Jagganathan:That's true for any technology. But, going to the earlier points, we wouldn't be talking about AI if people didn't get excited about labeling cats, dogs, horses 10 years back for example so I think that there is something to be said about really having it as an accessible technology where everybody can relate and be part of that technology. But name one technology which doesn't have a dark side this is true, maybe that's it.
Kimon Michaels:Maybe I'm a romantic. I throw out matchmaking. Do you really want it to be purely algorithmic driven? Oh, yes, yeah, thank you.
Francoise von Trapp:No, let's throw that right out the door, and that can be a little creepy too. I mean, we don't need that. Yeah, matchmaking, I would agree.
Kimon Michaels:And my wife probably wouldn't have picked me if she ran it through an algorithm.
Francoise von Trapp:Any others. That shouldn't be applications that we should not be investing in on the AI side.
Mark Kuemerle:I think a thing that we're just always going to have to be watchful for as a society is especially when we look at LLMs and generative AI. You can kind of train a model to do what you want it to do based on what you give it as inputs. So we do need to figure out how we can make sure that we have accurate input so that we're not building models which are maybe helping people by telling them incorrect information.
Kimon Michaels:It's a fine line between an homage and copyright infringement when it comes to art and writing, et cetera.
Francoise von Trapp:Well, I have run things through ChatGPT to see if the writing comes out better, but you always have to go through and edit them and make them personal, because, again, that's an area where having human interaction can improve what you get output from AI versus, maybe, the medical decisions. There are times where you use it as a tool to assist, not to take over. Okay, well, that's all we have time for, but I really, really appreciate your time. This is a great conversation and I look forward to having you on again.
Kimon Michaels:Enjoyed it. Thanks for having us.
Francoise von Trapp:Thank you very much. Thank you Next time on the 3D Insights podcast, recorded live at IMAPS DPC, we talk to officials and students from University of Arizona and Arizona State University about the importance of an industry-academia-government collaboration in building a solid semiconductor ecosystem in the US. There's lots more to come, so tune in next time to the 3D Insights Podcast. The 3D Insights Podcast is a production of 3D Insights LLC.