Judy Warner (00:00.992) Hi, Anrock. Thanks so much for joining us today on the podcast. I'm really excited to learn from you about 3DRF modules. So before we get started, can you tell us a little bit about your technical background and your role at Keysight? Anurag Bhargava (00:17.458) Hi Judy, yes, pleasure to be here. Thanks for inviting me. I hold a degree in electronics and communication engineering. I started way back in 1998 as a scientist at the Space Application Center, involved with satellite payload designs. In past many years, I was spent in Keysight, Agilent, and currently I'm based at Southern California, taking care of customers involved in cutting edge designs in RF and microwave. So here I am talking to Judy Warner (00:50.72) Well, I had a stint in RF and microwave PCBs. And so I have a keen interest in the technology. I think it's so fascinating. Of course, it's in everything right now. So I wanted to invite you in because I learned about some teaching you did regarding RF modules. So I thought our audience would like to learn about that. So why don't you start by talking to us about what exactly for definition purposes 3D RF module design is and what the applications Anurag Bhargava (01:28.496) Yeah, if you think about 3D RF module, it's a pretty niche technology. It's basically system as well as a typical RF front end, which people design now to reduce the form factor so that we can fit in more functionality into a small real estate. The applications could be pretty diverse, but the very common things you could see in our mobile phones, where there are so many bands, so many technologies now leading up to millimeter wave in 5G, FR2 frequency ranges. So all the way from 2G to 5G, everything is integrated into a small form factor. So on its own, it's a system in itself containing various kind of filters, amplifiers, switches, and so So it's a pretty, you know, upcoming technology and I would say it's increasing at a very rapid pace these days. So it's pretty interesting field to be Judy Warner (02:36.256) So is it a packaging solution then? Is the module like an IC package where you're integrating multiple functionalities into a chip level package? Anurag Bhargava (02:50.982) Yeah, you could say that it's largely a packaging system. But as I said, it's system on its own, but largely they revolve around packing different heterogeneous technology into one kind of module. So it could be like multi -layer laminate, packing different kind of technologies if you think about it. Because each technology has its own advantage and disadvantage. So when you think about. a larger system, for example, something for switch, the switch between different bands that requires a particular process, which is more suitable for such circuits. Similarly, if you think about power amplifier, which is needed to transmit our signal out of communication device, it has its own kind of technology. And the whole thing about 3D RF module is how to put together or how to integrate some of those technologies into a common interconnecting kind of platform. So in a way, yes, it's a packaging technology and lot relies on how you orient them, how you assemble them in a minimum possible space to provide the required functionalities. Judy Warner (04:10.486) So you mentioned, of course, that it's a heterogeneous system and package. So for our audience sake, why don't you share a little bit about what kind of problems or challenges engineers are facing? Because I imagine, like you said, the integrations of dissimilar parts is going to be one of those challenges. So can you unpack that a little bit for our audience? Anurag Bhargava (04:39.418) Yeah, sure. mean, if you think about like any, you know, multiple, you know, technology integration, you know, having a properly working 3D module, it's no small task. It takes a lot of effort and a very good design practice to come up with a successful design. Now, if you think about, you know, the individual technology itself, for example, somebody designing a switch, or a power amplifier or a low noise amplifier. Just to meet the stringent requirements of today's communication world itself is a challenge, where you need to produce an IC which meets all the standard as well as keep the cost down. Now, you double it up when you try to integrate those individual technologies to work together in a very small form factor where Everything is placed so close to each other that interference from each other can also play a small sport. And traditionally, if you think about packaging is always the roadblock of how fast technology can progress. ICs are going to two nanometers or less. That means we are able to produce a very, very fast rise and fall times, which extends the frequency space. But packaging. largely has been the main bottleneck of why people are still struggling to make some of these things at higher frequency. And the root cause is the parasitics caused by some of those packaging technologies. And as we all know, higher the frequency, even the smallest artifact on your design will play a very, important role because everything is very appreciable in terms of wavelength. So even a solder bot on your integrated module can be of electrical significance, which we don't see at a very small frequency. So some of the challenges are basically the ones which designers are dealing with. So it's a multi -fold challenge. Number one, when you conceptualize such product, how do you ensure a system -level performance? Anurag Bhargava (06:57.788) How do you do specification partitioning, what you want from each of those blocks? Then you go on to design those blocks. That's the second level challenge. And once individual teams have designed their own circuits, whether it is on a schematic level or layout level, how do you ensure that all those designs work together as a system and you are able to simulate? and characterize it with stringent wireless waveforms, which is basically the modulated standards we are dealing with, like 5G and now upcoming on 6G and so on. So once you cross that hurdle of optimizing and simulating all your circuit level designs together, then you come to a challenge of how to find a layout tool which can allow you to do the complete heterogeneous layout design. And once you are able to cross that hurdle, then it comes to really verify that 3D electromagnetic performance, which people typically do with the field solvers, to make sure there is no interference, there is no cross -coupling between different densely packed ICs or components in a very small space. And finally, more and more things you integrate. We are moving at higher power levels. People are looking to extract as much power as you can for effective communication. Then thermal becomes another challenge, because more things packed in a very dense space. That means the temperature effect of one component or IC will likely affect another one. So performing multi -physics kind of analysis in terms of electro thermal becomes another challenge. Because if you don't take care of the thermal, it will directly affect the reliability of your product. The performance may not be proper, or the lifecycle of the product might be compromised. Because remember the thumb rule in electronics, every 10 degrees rise in temperature will cut down your life by half. Anurag Bhargava (09:18.468) So the temperature management. Judy Warner (09:18.902) Wow, that's statistic I've never heard but a good one. So 10 degrees, you're saying centigrade or Fahrenheit. Anurag Bhargava (09:27.782) Mm -hmm. Centigrade, so 10 degrees centigrade roughly cuts the product life by half. So designers have to be concerned with making sure that the temperature or the thermal is very, very well characterized and defined. And we need to work out the proper dissipation mechanism so that the whole system can cool off as quickly as you can. So there are multi -dimensional challenges to designing such a product. And the need of the hour is to find a proper ecosystem where different technologies can be analyzed to the best possible extent. And that's what you will see nowadays, different tool vendors, different companies are trying to solve problem in their own aspect. is providing a very interesting, I would say, ecosystem for designers to work Judy Warner (10:36.534) Well, I have so many questions and as you were talking, I was thinking a lot about this ecosystem. First, as far as tools, of course, Keysight is sort of de facto high performance simulation, but you mentioned the design tools for the 3D module itself. What sort of design tools are being used for that, first of all? And then I want to ask you several more questions. Anurag Bhargava (10:39.612) Okay. Anurag Bhargava (11:04.028) Sure, as I said, this is a heterogeneous system on its own. So that simply means there is no one tool which can do all different aspects of such a system design. So you would need tools which has capability to do, let's say, design, which some people like to call it pre -layout kind of design. Judy Warner (11:29.558) Mm -hmm. Mm -hmm. Anurag Bhargava (11:31.25) And that tool should have capabilities to perform, you know, from board level aspect to the IC level aspect. And then finally, when you cross the turtle, as I previously talked about, you need a tool which can perform, you know, primary electromagnetic simulation. It could be what we call as 3D planar, like for IC level or a PCB or a module level. Or it could be a kind of mix of both, containing both 3D RF, which is full wave 3D electromagnetic tools, or with a combination with cleaner. Because different technologies serve different markets where they are more efficient while maintaining the accuracy. So again, and finally, when you start doing layouting where you combine your package along with your IC, Judy Warner (12:19.968) Mm -hmm. Anurag Bhargava (12:30.61) That's another challenge because you're dealing with, let's say, a nanometer precision up to the millimeter or centimeter precision. So those are some of the interesting things and different tools. And in ideal ecosystem, if your tools are interoperable, whereby you can leverage the design from one domain to another, that's possibly the best designer can ask for. Otherwise, it's always Judy Warner (12:50.003) huh. Anurag Bhargava (12:59.524) throw over the wall kind of concept where one team is able to do something and they just throw over the wall and then let other people worry about rest of the stuff. That's not Judy Warner (13:02.035) Yeah. Judy Warner (13:09.704) And then you throw it back and forth. Well, it's really interesting that you bring this up because probably over the last year or so, I've been tracking this trend towards ecosystems, which is sort of why I named my podcast, The Ecosystem, because I saw that, for instance, when I was with Altium is creating API. So it did have the interoperability that you discussed. And I know for instance, that Keysight's working with ANSYS, which probably gives you some of that simulation and probably other things. sort of, it seems to me, but I'm interested in your thoughts on that. People are saying that it's so complex and so fast, or the frequencies are so demanding that we're being forced into that place. We're being forced out of silos because it just doesn't work without it, or it's not efficient. enough is in the case of 3D modules is that something you're seeing as Anurag Bhargava (14:15.143) Yes, absolutely. And you're correct in saying Keysight. And there are many other vendors as well. So at Keysight, we understand this challenge. We understand what it takes to come up with a successful product. Because unlike some other EDA companies, I mean, if you look at the Keysight's history, we have been a hardware design house. Judy Warner (14:21.706) Mm -hmm. Anurag Bhargava (14:39.814) So we designed some of those very, high frequency stuff ourselves, which goes into our instrument. So we really understand what it takes to come up with a high performance module or a system. So recently, in the recent past, looking at the trends where technology is growing, we are collaborating with other industry leaders, such as Ansys, Synopsys. We have been working with vendors like Cadence for many years. So we understand and recently we also looked at LTM and that's where we are also trying to address that particular design segment. So the need of the hour is to have a kind of open ecosystem as much as possible, which is interoperable. It gives a higher return to designers and their CAD teams to make sure that their design flow is as cohesive as it can be. Because as I said in the very beginning, no one vendor can address multiple, multiple challenges because it is a very niche area, very specialized area where you need different, you know, know, bearings to actually come together for a smooth ride in terms of design flow. Judy Warner (16:00.086) I think there's going to be some interesting learnings as the technology drives us to create this interoperability in different domains like this one that are very niche. it's just going, it's an interesting time, I think, because I think there'll be some positive unexpected consequences I think that we'll discover. Cause the silo thing has been around ever since I have, and I've been in this industry for 30 years. And it's always been a problem, but we could work around it. And I think for these really high performance electronics, we can't do it anymore. I, anyways, I'm encouraged by it. And hopefully in the work you're doing, you are as well. Yeah. And regards to the, the simulation you mentioned, particularly when it comes to RF, are going to have antennas, right? And then there's the things that become antennas, but not on purpose. And so how do you, I mean, it almost seems impossible to me in some ways. And how are you able to navigate that piece? Anurag Bhargava (17:06.096) That's true. Anurag Bhargava (17:16.804) Yeah, I mean, I like your word. And believe me, every RF system anybody makes, it's either a desired antenna or an undesired antenna. And that is why we have things like EMI EMC, right? And the whole concept of EMI EMC is there is something unintentional going on if you don't take care of the fundamentals. Judy Warner (17:29.859) There it is. Anurag Bhargava (17:45.094) you know, from the very start. And that is why, you know, apart from the regular analysis, which we do to do such designs, it's very important to simulate the product in a near fabricated, you know, kind of situation where things are where they are supposed to be and everything working together as it's supposed to be in real Judy Warner (18:04.416) Mm -hmm. Anurag Bhargava (18:12.87) so that we can take care of some of these cross couplings or radiations getting coupled from one to another. So that's a pretty interesting and a very challenging problem to solve. But you can only achieve this once you can assemble your design in a manner where you have all those different heterogeneous pieces assembled together. So one of the challenge at Keysight we are trying to solve is to provide that kind of infrastructure, an open ecosystem where designers, no matter what tools they are using in their own respective designs, can still bring some of those technology pieces and assemble it the way a real product might be assembled. And then with the rapid improvement in the 3D simulation technologies, We like to call it multi -technology design. I don't know whether it's a industry standard term or not, but at Keysight, you will hear a lot of things about multi -technology. And what multi -technology essentially means is anything which has more than one kind of process being assembled together. It could be as simple as, you know, package and IC, but IC mounted on a PCB. Now, you could Judy Warner (19:35.733) Mm -hmm. Anurag Bhargava (19:37.326) multiple ICs mounted on a package or a PCB. So as soon as you start integrating those different heterogeneous pieces together, it becomes a multi -technology. And to assemble that and to simulate that in the best possible efficiency are the key to guarantee you the required outcome in a more predictable manner so that you can anticipate and debug or find the issues very early in the design cycle rather than waiting too late and testing some prototype in your lab and then you discover such problems, then it's very hard to fix because it usually means redesigning something. It could be IC, it could be package, it could be something else. And that takes a toll on your time to market, which is such a highly competitive word. Even if you can reduce one tape out, It saves a lot of money for company and improve their time to market or objectives. And another interesting thing if you think about is how do you put all this together in the most efficient manner? Because you can always do manual things. But anything you're doing manually, it's not really long lasting because that means it's still very heavily manpower dependent. And with the rapid pace, it's need of the hour that you put around certain automations behind the scene. And individual or CAD teams in a particular organization have a more automated streamlined flow. And that's one another thing which you will see in the industry today is more and more people are talking about automation coming up with a strategy where you could do things in a more efficient manner. And we understand that. And similarly, at least I can say from Keysight perspective, we are putting a lot of effort in coming up with Python APIs for our product, because Python is one kind of widely accepted language. There's an abundance of manpower who are Python skilled. The new graduates coming out of college are Python aware. So that is proving to be a pretty Anurag Bhargava (22:02.586) pretty nice, I would say, And most of the design tools, especially RFMicrowave, if you think about, they are already kind of supporting Python in one way or another. So that can act as a very nice clue to bring multiple tools together and have an interoperable kind of ecosystem. And that's what we are doing from our side as well. Putting effort on automating some of these, yes, you could do it with Python. Making tools friendly with Python gives all the handles and levers for cat teams to utilize them with those. So that's the way I see technology going forward without automation. It's kind of very difficult to conquer those upcoming challenges. And just imagine what SIGSRI is going to bring us. Judy Warner (22:59.254) Lord help us all. And not only that, I hear people talking about on the high speed digital side, 448 and I'm like, yeah, we haven't even, we're still struggling with 112 to 24. We're always running so far ahead of what we can execute on, but it's a very interesting time. Is there anything, I mean, this seems like such a deep, rich conversation and technology and enabler. Anurag Bhargava (23:10.674) Shoo shoo shoo shoo Judy Warner (23:27.956) Are there any things for our audience that you could offer like just as a workflow solution? One thing they might start doing right away that can help them be more successful in this area, maybe a bad habit you see in context of the customers that you work with that are hardware engineers that you can suggest. And I'm going to give you a two -part question here. So what's like, one thing, one or two things they can start applying today. But I also want you to talk a little bit, which is what drew me to you was this, I think it's a webinar that you did on RF3D modules, which is what sort of attracted me to you. So I'd like to know what can they do now, and then talk to them about what's in that webinar and how they might, when they have more time, go a little bit deeper. know what they would find in that webinar specifically. Anurag Bhargava (24:32.688) Yeah, sure. I I will try to answer that question, but it could be very subjective because, you know, different customers like to do things differently. Different design teams approach the same problem in a very different manner. So one thing which I can say people can do right away is to, you know, take a step back, look at their design process, what they do right now. One to two to three to four. Judy Warner (24:46.269) Of course. Anurag Bhargava (25:02.296) and eventually they get the product out. Try to analyze which step is taking their most amount of time and where they iterate quite a few back and forth. Because those, by looking at that, you could identify the bottlenecks or the key steps where you spend most of the time. and try to look around and see if there are better solutions available or better ways of doing things available. Because sometimes it's very easy to get so involved in something which is working right now that you stop looking outside. There are so many things happening around the industry that you might be able to find a very, let's say, appropriate solution which can help you reduce some of the time you're spending in doing some of these things. Anything which you can automate, which you are doing manually, we're always going to give you time advantage. And as you said, the second part, some of those challenges I try to cover in my webinar topic, which is focused around integrated workflow for 3D RF module design and simulation. While due to limitation of time, even in that webinar, I couldn't spend enough time. over each aspect, but at least by looking at the webinar, it's just 40, 45 minutes. So anybody can watch that. At least it shows you how to address some of those key challenges and what solutions you have available right now to go and deploy and apply. Even if not as a whole, you can still do incremental progress. Remember, any step you make, So it's going to give you certain improvements. And the whole stress, as I talked about, is having an open ecosystem where different tools can start working together, no matter if it is, let's say, for company X or company Y. We are committed to providing an ecosystem which is open, versatile. Anurag Bhargava (27:19.966) And customers can adopt the way they want to adopt it in their own design flow. So just looking back, analyzing where you're spending more time, how can you make it more efficient? And if you need more details, you could contact anybody at Keysight for that matter. Now here, talking about Keysight, where you can talk to your own vendors, no matter which one you are working with. If you don't find solution with them, talk to Keysight because we are working with various vendors creating and opening up such ecosystem for designers to take out of. And if you're not aware of Keysight tool, one thing I would like to mention outside my work, because RF microwave has been my passion for 25 plus years now, I maintain my own YouTube channel. where I provide not only tool related videos so that people can ramp up much quicker with the design tool. It also has quite a few real practical design tutorials, which anybody can refer to and understand the intricacies of RF Microwave Design. So some of those things are at everyone's disposal, it's free, there's no charge around it. And learning something new never hurts. So you never know when it's going to help Judy Warner (28:50.294) Well, I couldn't agree with you most and you beat me to the punch on your YouTube channel because that was the next thing I was going to have you discuss. So thank you for that. And for our listeners, I will put Anraag's YouTube channel as well as his webinar below so you can dig much deeper. I can only imagine how long you could talk about these things with the complexity of these systems that are packages all the complexities. So thank you so much for at least taking this short time to, you know, give us insight to what's happening in the industry and this interesting niche area. I think it's fascinating and, and I have many more questions as I'm sure our listeners do. So I will also be going and consuming those things. So Anrock, thank you so much for joining us today. I've really enjoyed our conversation. Anurag Bhargava (29:48.124) Yeah, thanks. for inviting me again and pleasure talking to you and audience and one departing thought I would say never lose any battle alone. So if you need any help from my side or from Keysight, we are committed to help you out. Please feel free to reach out to us and discuss your situation with us. I'm sure we'll find. some other solution to anything you are struggling with or you would like to improve upon. So with that, thank you very much for inviting me and listening to me. It's been pleasure talking to you. Thank Judy Warner (30:28.214) For our listeners, thanks so much for joining today. I hope you've enjoyed this conversation with Anurag Bhargava as much as I have. And I encourage you to go check out those links, both to his webinar and his YouTube channel, and just dig in if this is a technology you're looking at and thinking proactively about your workflow and just taking a step back and noticing what's happening in the industry. One final thought I have. is that some of the things that are being automated in the industry are taking those sort of mindless repetitive things and automating them, which gives you much more creative time to do. And it's happening across the industry as well as inside Keysight. So dig in and please sign up for the ecosystem newsletter. Whenever we find good resources for you, we'll be sure to put it there like this one. So thanks so much for joining us today. We'll see you next week. Until then, remember to always stay connected to the ecosystem. And that's a podcast.