EDA and the Shift-Left with Mike Ellow – Part 1 – Transcript
In our first conversation of the latest season of the Industry Forward Podcast, Dale and I were joined by Mike Ellow, CEO of Siemens EDA! Mike holds 30 years of executive sales and technical management experience in Electronic Design Automation (EDA). We invited Mike on to gain his insights on the transition to software-defined products and systems, the shift-left of engineering processes, and especially the increasing need for the integration of EDA and semiconductor development into the creation of complex systems like airplanes, cars, and more.
Listen to the episode and read a transcript of the conversation below!
Conor Peick: Mike, welcome into the show. Thank you so much for taking the time to chat with us today. I think I can speak a little bit for Dale here. Really excited that you’re joining us and we’re really looking forward to this. So you know, before we dive too deep into the discussion. I was wondering if you could breakdown just a few key aspects or characteristics of semiconductor development, EDA tools, just to set a baseline of understanding for me, maybe mostly me and any listeners who aren’t as familiar with the intricacies of EDA and semiconductor development?
Mike Ellow: OK. So First off, thank you for the invitation. My pleasure to be here. But as we take a look at EDA and semiconductors. So EDA stands for electronic design automation and if you take a look at what we do. In essence, the way I like to describe it is everyone typically has a smartphone. And the electronics that sit inside the smartphone. Are what we help the pose. Use Apple as an. Do all of the design work for across multiple teams that are geographically dispersed? We help them organize all of the information from specification into actually producing something that can then go to a semiconductor manufacturer to produce. That goes inside the cell phones themselves. You take a look at. What’s going on with AI and all of the hyperscalers that are out there with high performance computing that are used to train AI? We are the backbone as far as the silicon that sits inside all of the CPUs and GPUs that are being used for the model training you may. Have heard about. They’ve been in the news for the past year or so as arise as riding the. Explosion of AI with their GP US is foundational for the training of those models we work with. For instance, NVIDIA, to help them design those GPUs. We are very pervasive across all of the. I mean, we like to think that there is not a semiconductor out there that is, that has not been impacted somehow by the EDA technology that we develop.
Dale Tutt: Mike, again, thank you for being here today. And it’s really this is exciting for me. I think a lot of our listeners. I really come from a lot of different industries and their experience with electronics is it’s a box that you maybe buy from somebody else or if they’re doing some of their software themselves again, maybe they’re buying purpose-built chips or you know off the shelf. Semiconductors and so I’m excited about this today because you know really. You know, we sometimes think about the electronics and, you know, really is the brains of the products and it’s becoming more and more important today. So you know with. The big trend in so many industries, this transition towards software defined products and systems in your. More and more of the systems are being defined by software and not necessarily hardware. I think back to recent change. It needed to be made to one of my automobiles and there was a problem with it and I thought we’re going to have to take it in and have some hardware changed out and they downloaded a new software. To it, new software load and it. It fixed the car and so it. It was pretty amazing and really just changes the whole changes my life so. So with this big transition towards software defined products and system, what is a software defined product or system and what are the biggest challenges or pain points that are introduced with the growing emphasis on software as well as electronics?
Mike Ellow: Good question. So don’t be shy. I think when you coax me onto your podcast here, you said this was the most exciting conversation you were gonna have ever. Both both in the past and both in the future as far as where we’re headed.
Dale Tutt: Absolutely. And I’m looking forward to it. Absolutely.
Mike Ellow: Just so we’re on the same page here though. So you know this is not just the electronics piece. The electronics as you take a look at the hierarchy, sits at a level where you’re actually putting the semiconductor onto PCBs. The printed circuit boards and the printed circuit boards maybe. In a combination of printed circuit boards, in order to provide a certain level of functionality, which is the electronic delivery associated with. So we span. You know what happens all the way back into the architecture phase and why I wanted to replay that because we start taking a look at software defined products, right? Seen over the past three or four years, a movement that many of the larger. Product integrators. System integrators are looking at where their value comes from in the marketplace and how do they differentiate and is through software now, right? Used to be that. You know some level of the mechanical. Piece itself or the manufacturing process where the AIDS for some level of differentiation, but the world’s changed. And if we use the automotive example, especially around when you’re looking at autonomous driving. Now a person driving a car is no longer paying attention to the road. So what do they do? They’re in a enclosed environment where the car manufacturer now has to figure out ways in order to stimulate the census. In order to give them entertainment. Or an experience so that they are not bored because they have more time on their hands.
And as you take a look at now for a lot of these OEMs using the automotive example, again their brand will now be defined by what that user experience is inside that. Cockpit. Right, so this is a fundamental change going on out there and it’s not just sitting in automotive. You take a look at what happens in aerospace and defense. You take a look at what’s happening in heavy machinery associated with the move towards. More autonomy associated with. It is the software that now defines the differentiation and in a world like. You no longer can use the existing infrastructure in order to. Have predictability around schedule, functionality and cost of the delivery of these complex system of systems. And so the first piece, you’re really taking a look at is well, what am I going to do in software versus what am I going to do in hardware in this space? Is the. And then how does that influence all of the other swim lanes associated with the development of these complex system of systems and? In that early architecture exploration, as you’re taking a look at real or synthetic workloads running on virtual representations of the semiconductor, you’re doing a lot of experimentation in order to understand well what is the compute platform. That best optimizes the software and why this is such a dramatic changes in the past.
So you know the software teams were handed existing hardware platforms, existing silicon configuration and said build your software here. And what? Well, there was compromises made in software because the hardware was already fixed. If software is your key differentiator now, you can no longer live in a world where I compromise all my differentiation. I must have the ability to properly configure the entire system. System around that deliverable of the software and this is what is now an exciting opportunity. For companies out there. It’s a threat to some of the established companies also because now they have to adapt at a much faster pace with a legacy associated with not only how they do things but mindset in order to be competitive in the world of changing rapidly underneath them.
Dale Tutt: Yeah, I think that’s a really good. And you know, the whole notion of the software being the key differentiator that in the past, you know that cars, auto companies, aircraft companies, doesn’t matter. That they knew what they were doing. I had it figured out and they knew how to do the mechanical design, electrical system design, and now it’s like a new skill set and it’s that they also have to pick up with the software and electronics piece and really being able to do it. A multi domain way you can’t ignore. That this is all the elements have to come together in a way to really, you know, to make to make the system work and it’s more and more important I think that you know. When we made the transition to smartphones. And this whole notion of putting apps on the smartphone people were, I think at first were like, I don’t need all this stuff. Just need a phone and nowadays how often do you actually use it as a phone?
And the same thing is happening. I see it happening in aerospace and automotive everywhere else that at the you know, first glance there was you know a few years ago working on one of the programs that I was. We were struggling because we were really trying to adopt a mindset of we need apps. We need app to add this new functionality and for everybody. It was just it was a totally new muscle motion for them. That really is a big change. I think as companies are trying to adopt new way of doing business and really you know to meet the consumer demand of the future. It is a big change.
Mike Ellow: Well, you mentioned apps and if you take that to a logical next step in the conversation that we’re having. The integration of a. Stack and what the platform looks like versus what you enable as far as the apps. On top of that and how do you, you know, contemplate the expandability of your software platform over the lifecycle of an asset such that you’ve got the ability to add those apps of? Future, which you may or may not have any line of sight now. You’ve got. Comprehension based on how you do the architecture, exploration, and plan for expandability of the of the compute infrastructure to handle that over the plan lifecycle is also one of these areas. That I think we see more and more pressure on because we don’t design anything for a point in time. We design for a point in time only to get first article out. First article is yes, we’ve got a point in time where we’ve got to have the coming together of all of these complex domains with the software writing on top such that first article comes out and it hits what we expect on cost. Schedule an on price. Once it’s deployed, that’s a whole different story.
As far as what’s happening over the life cycle of this thing, and you take a look at where we’re headed on the semiconductor side and stuff, how we embed active monitors into the semiconductor such that we understand what’s happening to that in real world operating conditions, even though we will have planned and provided margin in that design phase for what we anticipate those operating conditions to look like, maintaining lifecycle, things in the field, especially when you start radically updating your software stack based on the explosion of functionality that people want to experience over the life cycle of their asset, adds a level of complexity where we need to be fine-tuned into what’s happening on a very frequent basis with the asset in operation: to be able to extract performance criteria, aging information, reliability information and the ability to model the increased software stack in virtual views because this will be not only you know, a hardware platform you have that you can do the extraction, run your software stack on before you ship your over the air updates. Which, I’m curious as to why an over the air update in your car changed your life. I mean that is really a bold statement you made there, but makes me wonder what your life’s really like. But it’s there’s a lot more.
Dale Tutt: Man, I didn’t have to take the car into the shop. That’s what changed my life! Car wasn’t gone for a week!
Mike Ellow: Oh. My God. Yeah! But the world in the pace at which this is evolving is, is crazy, right and not only you know, when we talk about this it’s kind of a sterile conversation because this is an ecosystem that is producing these system of systems. So how do we interlock all of the different companies that have to be aware on a real time digital backbone? To understand during the, and I have a way I view this, on the design, optimization, verification, implementation, manufacture, deployment, and maintenance of these complex system-of-systems over time. It’s a daunting, daunting problem, but, what I think is exciting is you start taking a look at where Siemens sits and all of the assets we have. We’ve got a very interesting lineup of capabilities that as we work together now to get those things operating more efficiently together and take a look at the expansion possibilities of the capabilities in order to map against this big problem. It’s I think we are one of the few companies in the world that are our customer base can look to and say, “You know what? Siemens, you’ve got a lot of promise. You just need to deliver.”
Siemens Digital Industries Software helps organizations of all sizes digitally transform using software, hardware and services from the Siemens Xcelerator business platform. Siemens’ software and the comprehensive digital twin enable companies to optimize their design, engineering and manufacturing processes to turn today’s ideas into the sustainable products of the future. From chips to entire systems, from product to process, across all industries. Siemens Digital Industries Software – Accelerating transformation.
