Francis Chow of Red Hat

Overview:

JOHN: Today in The Garage, live from CES 2025, Red Hat. Let’s go!

JOHN: Welcome to The Garage. I’m John Heinlein, Chief Marketing Officer from Sonatus. We’re recording live at CES 2025 from Las Vegas. We’re pleased today to welcome Francis Chow. Francis is VP and General Manager of In-Vehicle OS and Edge at Red Hat. Francis, welcome to The Garage.

FRANCIS: John, thank you so much for having me.

JOHN: You know, we’ve known each other for years. In fact, we met during Covid, as I recall.

FRANCIS: Yes.

JOHN: And our first many, many meetings were virtual. So, it was so nice to finally meet you in person a few years ago, and we’re thrilled to have you on the podcast. Can we start by telling us a little bit about you and your background?

Meet Francis Chow

FRANCIS: I am an engineer by training. I actually started my career in semiconductor, and I’ve gone through a bunch of roles.

JOHN: So did I.

FRANCIS: From design to corporate development and then marketing, pre-sales, and towards the end I was managing the telecom business for a semi company. And then, I ventured into the software world. So, in the past nine years, I have been in enterprise software.

About Red Hat

JOHN: Fantastic. And tell us about Red Hat in general and your role at Red Hat.

FRANCIS: Red Hat, for those who are not familiar with us, we are the largest open-source software company. And we started our journey about 30 years ago when we decided that, instead of filing patents for expensive proprietary products, the best way to design software is with open collaboration. And my role at Red Hat, as you said, I’m responsible for in-vehicle OS and edge, right? That sounds a bit strange, partly because I have a dual role.

JOHN: Right.

FRANCIS: Part of that is a product role in the business unit where we’re building and trying to grow the business with in-vehicle operating systems. And we can talk more about that. And I also have the portfolio responsibility for edge. So, trying to work through our portfolio on that and applying those on edge and operations technology use cases.

JOHN: And Red hat was acquired by IBM. Was it last year? The year before? Is that right?

FRANCIS: Red Hat, I believe, was acquired by IBM in 2019 or ’18, in that time frame.

JOHN: Okay, so a little further back. How has that synergy been helpful for you?

FRANCIS: It’s been going great. I think IBM graciously gave us a lot of freedom on how to run our business. And there are a lot of collaborations between the two companies today, both on technology and go-to-market insights.

Fun Fact

JOHN That’s great. So, we always like to get to know our guests right at the beginning. And I’d love for you to share with us a fun fact about you.

FRANCIS: Sure. I think a fun fact about me is, at one point in my life, I was in a Catholic band.

JOHN: Catholic and band are not two words you hear together all the time.

FRANCIS: You’re absolutely right. So, the background of that was that I’m a born Catholic.

JOHN: Me too.

FRANCIS: I’ve been Catholic all my life, and I’m also into music. I was doing choir, choral music for about ten years, and at one point we thought, hey, you know, the music for mass is just a little bit boring, maybe we should spice it up a little bit. So, that was kind of the motivation for us to form a band. I do vocals, I play guitar, and it’s been a fun journey.

JOHN: That’s fantastic. I always try to share a fun fact back to match my guest. And so, I think I have to share back that I played classical guitar as a kid for many, many years. I’m very rusty and I’m definitely not as talented as you, I’m sure, but it was something from my background. But I also need to share a bonus fun fact, and I don’t even know if you know this. But you were talking about your background with Red Hat. My previous start-up company, many, many years ago was a company called Transmeta, which was a…

FRANCIS: I went and talked to Transmeta when I was in semiconductor.

JOHN: That’s right. And, three people before me — I was like, number 88 at Transmeta — and three people before me was a guy named Linus Torvalds.

FRANCIS: No kidding?

JOHN: He was my coworker for many years. He was hired like a month before me. His first Thanksgiving in the US, we sat together at Thanksgiving for his first Thanksgiving in the US. Fantastic guy. He’s a really clever guy. I really got to know him. And so, when Red Hat came out and went public and everything, it was really exciting to see his work getting, you know, made public and more common. It was great. So, there’s a bonus fun fact for the day.

FRANCIS: That’s awesome. That’s awesome.

Why is SDV Important?

JOHN: So, in the podcast here and also at the show, we talk extensively about SDV. I think it would be good to start by saying, you know, why is SDV important? Why should we be talking about SDV? What are the opportunities and challenges that SDV brings?

FRANCIS: I think if you look at the automotive industry, what we’ve been observing is that the customers are demanding more capabilities, more safety, more functions, more personalization, and…

JOHN: The customer meaning the driver.

FRANCIS: The driver. Yes.

JOHN: That’s right.

FRANCIS: And the way to do that is not to do it in the old way that the industry has been doing it. Which is building bespoke systems that are very difficult to reuse and maintain and update over time. We believe there is a way to bring from other industries that have gone through the journey of software-defined architecture. Like in the data center.

JOHN: Exactly.

FRANCIS: Like in the telecom network.

JOHN: Right.

FRANCIS: And a lot of these learnings have been proven in other industries. And we can bring that know-how and expertise into automotive. And what we think is, because of the complexity of the vehicle’s systems have been becoming. There is an opportunity for us to simplify the design, maintenance, and update. And the way to do it is to leverage a lot of the software-defined concepts that have been proven in other industries.

JOHN: I think it’s such an important thing you’re saying there, because — we talk a lot about this in the podcast and also at Sonatus — because sometimes I get the pushback from customers or people of, oh gosh, software is hard. Or, oh my gosh, you’re going to put software in vehicles. There’s already software in vehicles, but it’s done in a way that’s not maintainable, not scalable, impossible to validate. And many of these, as you said, many of these problems are known problems, are solved problems in software-defined networking, software-defined data center. Workload isolation is sort of my pet, favorite example. This is a solved problem in a data center where you’re constantly running next to a workload that you have no idea is there, and it doesn’t touch you. That’s the same capabilities we need to put in vehicles. Okay, now in vehicles, sometimes they’re real-time, sometimes they’re not real-time. So, there’s some additional complexities. There’s some additional safety. But the fundamental problem is well understood. And so, I think we, the industry, we, you and I, we’re trying to do the kind of innovation to bring that to SDV.

FRANCIS: Absolutely. I think there is a lot of learning that we can take from other industries to automotive. We’ve seen a lot of resistance. Partly, it’s because the way it has been done has been for a very long time. It’s proven but not optimal. A lot of advancements in software in the past 20 years have not been brought to the automotive world.

JOHN: Right.

FRANCIS: And what we hope is that we can be a part of the ecosystem to bring that into this industry.

How to Make SDV Profitable

JOHN: That’s right. And the other thing that we talk about — and we’re showing some examples of this in the booth, which is just right outside there today — is, customers are always asking, how do I make money with SDV? Okay, SDV is fine. I realize I need it. How do we make money? Well, the first thing is, I think it’s going to bring cost reductions. It’s going to bring savings in the way things get engineered. The fallacy, the mistaken impression is that software will be more expensive. What they’re doing today is expensive. This will bring the opportunity — after a period of transition, after a period of learning — of a much better architected, much more scalable, much more maintainable thing that can save money. But then there’s opportunities for revenue along the way, and we’re showing some examples of how. Providing features to customers, providing capabilities, as you mentioned earlier, that evolve over time. The example — I was talking earlier to a customer out here in the booth — is a driver, they don’t get upset for paying for something they value.

FRANCIS: Yup.

JOHN: They get upset for paying for something they don’t value.

FRANCIS: Yup.

JOHN: And so our job in the industry is to enable the OEMs, enable the Tier 1s to provide valuable things that customers are excited to pay for. Please let me pay for that because that is such a great feature. That’s the promise of SDV.

FRANCIS: Absolutely. Like you said, SDV absolutely would be a way to address some of the cost challenges the industry is facing today. That’s right, like you said, right? There will be a transition that’s needed. OEMs and the ecosystems, we need to work together to figure out how to accelerate that transition as soon as possible so that we are not adding cost to the existing programs, rather than trying to figure out how to jumpstart a new paradigm of design. And I think, as you said, the most exciting thing, to me, is when you have a software-defined platform, the possibilities of having an ecosystem, writing applications, adding capabilities to that platform. Just like how we’ve experienced in your cell phone in the past 15 years. It’s just… who would have thought of what you can do with your cell phone today, 15 years ago?

Red Hat In-Vehicle OS

JOHN: That’s right. Exactly, exactly, absolutely. So, you run — among other things — Red Hat in-vehicle OS. I’d love for you to tell us a little bit about that offering and introduce that to our guests.

FRANCIS: So, one of the challenges that we have observed in the industry is that there’s no easy way to design software. Currently, a lot of systems are running on traditional RTOS. There’s a lot of hardware dependencies, at times.

JOHN: With things like AUTOSARs, of course.

FRANCIS: Very difficult to update and maintain over a long period of time. And so, as we move to a platform approach, silicons are getting more powerful, and there needs to be an operating system that sits between the applications and the hardware to give you that linkage, to provide the capabilities and make it accessible to upper-layer software. And provide that abstraction so that you, as an application developer, you don’t have to worry about the hardware dependencies as the hardware evolves over time. So, what we would like to offer is to, again, bring from what we’ve learned from data center and other industries, an operating system that is at the cutting edge. That has all the advancements in software in the past 20, 30 years to make it easier for developers to develop applications for vehicles that are still safe.

JOHN: Right.

FRANCIS: But the key thing is, we’re not compromising on safety.

JOHN: Right.

FRANCIS: We want to bring productivity, cost savings, while maintaining safety.

JOHN: That’s right.

FRANCIS: So, what we’ve tried to bring is a safety-certified Linux operating system from the enterprise world to the automotive world.

JOHN: That’s great. And I mean, obviously there’s no criticism of RTOS vendors. There’s a lot of important RTOSes out there for many years. But what you said that’s important is, it’s a modern operating system that’s easy for others to develop on. And that’s not typically the case with RTOS. It takes some specialized expertise to develop on an RTOS. You don’t have the same infrastructure that you have on a Linux-based, POSIX-based operating system. It unlocks a lot of productivity and a lot of scalability, I think, that’s not typically possible.

FRANCIS: And, that’s absolutely right. It’s just the amount of talent, for example, you can find with Linux expertise is probably 100 to 1000 times more…

JOHN: Yes.

FRANCIS: Compared to more traditional RTOS. And then, we talked a bit about AI earlier, right? Almost all of the AI technologies are developed on Linux, right? This is the right platform to develop new technologies.

JOHN: That’s right.

FRANCIS: And as we are getting into the AI era as we speak, I think there are a lot of even unseen potentials that we can achieve with the right platform.

JOHN: That’s right. I showed you some demonstrations earlier in the booth. This week at the show, we’re rolling out a lot of AI technologies across many of our products, and many more of them to come in 2025. All of that’s based on top of a Linux-based environment. So, we’re very excited to work with you.

Safety Certification

JOHN: So, you’re also going through your own safety certification journey. And I know that, especially for an operating system like you, it’s an especially arduous journey. Tell us a little bit about the process of going through safety certification.

FRANCIS: Yes, John, I would say this is probably one of the toughest things I’ve ever done in my career. It’s not like people haven’t tried certifying Linux for safety. In fact, there have been a number of efforts…

JOHN: Of course.

FRANCIS: …that were attempted in the past ten years by OEMs, by suppliers, so far without much success. So, what we’re trying to do is to bring in a new approach that is still compliant to the standard, right? So, we will be certified on ISO 26262 ASIL-B. But we’re taking a more tailored approach so that we have a way to certify pre-existing open-source software and still achieve the same risk objectives of ISO 26262.

JOHN: Interesting. It’s tricky with pre-existing software because a lot of the safety methodology requires going into the development methodology from day one to do it right. So it’s a different challenge, I think.

FRANCIS: It is definitely a different challenge. But in the end, safety is quality.

JOHN: Of course.

FRANCIS: And Linux is a well-proven, highest-quality operating system that supports many servers and applications today. So, we’re trying to kind of bring that proof point, right? To the industry that, hey, there are ways to prove that it’s also safe with our top-quality record.

JOHN: You know, Sonatus just completed our own safety journey because, just in December, we announced that our Automator product has a safety module that’s now ASIL-D (for delta) certified. Which is so important because customers want to use our Automator product in vehicles, but they’ve been asking us, well, is this going to compromise safety? Can I use this while the vehicle is in motion without compromise? And so, we’ve proven that that’s the case. So, we got to experience our own safety journey. It took us pretty much the better part of 2024 to develop the, you know, the capabilities, the infrastructure, the know-how, and to develop this custom module called the Automator Safety Interlock. And as you say, for you, every product has a slightly different approach. For us, it’s a safety module that sits between our product and the vehicle that’s a monitor that says, is this or is this not safe to do in this environment. And so, it protects and allows us to preserve vehicle safety. And also, similar to your situation, we didn’t have to prove or try to prove the entire Automator product because it’s a larger product. We were able to certify this more focused module that serves as a gatekeeper to ensure the actions are safe. And so, it’s easy for OEMs to understand. It’s easy for OEMs to be confident that it’s a great solution. So, it’s been a journey for us too.

FRANCIS: Yes. And to share a bit more about our journey. Like you asked earlier, we have mapped out our path to safety about a year and a half ago. We’ve gone through the first step in the middle of last year, which is the math library certification. And we got the certificate. And it was more like a proof point that our data approach actually works within the confines of ISO 26262. And I’m very happy to share with the audience that we just achieved the safety certification for mixed criticality.

JOHN: That’s great.

FRANCIS: What that means is, we certified a number of key Linux subsystems to enable mixed criticality. That means we’re able to run both safety and non-safety applications on the same OS without the need of hypervisors or multiple operating systems that’s been a need in the past.

JOHN: That’s great.

FRANCIS: And that brings dramatic improvements on software system architecture, resource utilization, design, testing, deployment without compromising any safety concerns.

JOHN: That’s fantastic.

FRANCIS: And we’re excited that, with this milestone, we are still on track to get a full product certification within this year.

JOHN: That’s great. Congratulations. We look forward to hearing that. Stay in touch and we’ll talk more about that when you achieve that milestone. That’s great.

Red Hat and Sonatus Collaborations

JOHN: So, we’ve been working together throughout this past year, Red Hat and Sonatus. We’ve been working together on some projects. I thought it’d be good to talk a little bit about the collaboration we did. Do you want to share some of your thoughts?

FRANCIS: Absolutely. And we’re grateful to have Sonatus as part of our ecosystem. And we’ve been great partners. I think we share a lot of common principles on how we bring software-defined vehicles to the world. And what we’ve been working on is to pre-integrate our products so that they’re ready to use, ready to deploy. A lot of the components that Sonatus has, have a component in the vehicle as well as a component in the cloud. Now, what Red Hat also brings to the automotive industry is not just our expertise in Linux, but our expertise in running cloud-native environments.

JOHN: Exactly.

FRANCIS: So, we can have the same operating system running in the car as well as in the cloud. So, that collaboration with Sonatus to have a pre-integrated solution on multiple use cases has just been great.

JOHN: That’s fantastic. We’re showing an example here at the show about how one can use our Sonatus Collector product for better data collection. We’re showing an example for battery management, but that’s just an example. It can be applied to many other subsystems. And I was just showing you earlier in the booth, that same Collector product doing many other capabilities with our new generative AI interface to Collector, which is very powerful. So, I think the sky’s the limit with what we can do with that collaboration together.

FRANCIS: Absolutely.

The Industry Big Picture

JOHN: Well, let’s think about big picture. If we pull back and think about the industry more generally, what’s your advice? And we have a wide range of listeners. What’s your advice to the industry and your outlook on where SDV is today, where it’s going, and what are the challenges we need to solve?

FRANCIS: What we are seeing is the industry is currently undergoing some economic challenges.

JOHN: Yeah.

FRANCIS: But this is not the time to pull back because technology is a way to address this challenge, and we believe it’s probably the best way to address this challenge. We’ve been talking about software-defined vehicle for years, if not decades. I think the way to do it is similar, actually, to how open-source works.

JOHN: Yeah.

FRANCIS: Let’s start by doing. We’ll contribute code in the open. We’ll work together with the ecosystem in the open. We’ll start by implementing things, executing things, trying things instead of talking about things.

JOHN: Yeah. You know, it’s such an important thing you say, and we often have this conversation with customers. They say, oh, it’s too hard, it’s too big. And I always say these funny anecdotes, and so maybe you’ll think it’s funny or maybe you’ll think it’s useful. But there’s an old saying it’s, how do you eat an elephant? And the answer is one bite at a time. You don’t have to swallow the whole elephant.

FRANCIS: That’s right.

JOHN: It’s a big problem. We can find parts of the problem to solve. And SDV is not a bit-flip. You don’t flip a bit and now it wasn’t SDV and now it is an SDV. It’s a journey of pushing SDV throughout the vehicle. We can deliver tons of value by peeling off parts that were inefficient, that are designed stupidly in the old way, and make them better, make them more modern, and journey, journey, journey towards a better, you know, of eating the whole elephant. But it can happen over time.

FRANCIS: And I would add that, that journey has to be done through collaboration.

JOHN: That’s right.

FRANCIS: Just like between Sonatus and Red Hat, no company can or should do this alone.

JOHN: That’s right.

FRANCIS: And again, we believe collaborating in the open is the best way to make it happen.

JOHN: Absolutely. Listen Francis, it is such a joy to have you on the show. I’ve been looking forward to having you here for a while. So glad we were able to make this work here at CES. We appreciate your partnership. We appreciate all the work you’ve done, and we look forward to great things ahead.

FRANCIS: Thank you, John. I really appreciate the opportunity to discuss this.

JOHN: Thank you. If you like what you’re seeing on the show, please like and subscribe to hear more episodes from us. We’ve got many more episodes coming from CES 2025 and more to come from our studio back at home. Thank you for joining us. We’ll see you again soon.