BlackBerry Limited (BB) RBC Capital Markets Technology, Internet, Media and Telecommunications Conference 2022 (Transcript)
BlackBerry Limited (NYSE:BB) RBC Capital Markets Technology, Internet, Media and Telecommunications Conference 2022 November 15, 2022 9:25 AM ET
Company Participants
Mattias Eriksson – President, BlackBerry IoT
Conference Call Participants
Paul Treiber – RBC Capital Markets
Paul Treiber
Right so and thank you, everyone for joining me it’s nice to be back in person. My name is Paul Treiber, I cover Canadian Technology stocks at RBC. I’m pleased to be joined by our – presenting our next guests from BlackBerry. So Mattias Eriksson who is the President of BlackBerry’s IoT Business Unit; and also Tim Foote, Investor Relations. So just you know, obviously, I think everyone knows Blackberry.
Question-and-Answer Session
Q – Paul Treiber
Mattias, so the first question to start is you joined BlackBerry about 18 months ago, you’ve been in the automotive software space through your technologies, what attracted you to BlackBerry, and specifically the IoT unit?
Mattias Eriksson
There is a couple of different things. First of all, I think in terms of the assets and the IP and the capabilities that the BlackBerry IoT business unit has in this space, very important to me difficult to replicate. And the foundation for a great position is an engineering foundation for building a great business. Secondly, Blackberry has a very unique position in this ecosystem, if I look at the customer base.
And the partner base, in a long lifecycle business like this, joint innovation with customers and partners over multiple years is critical. And unless you work with the best, it’s very difficult to gain traction, and Blackberry has a unique relationship with many of these partners. And then last but not least, the foundation of our business has multiple structural trends that are pushing both growing the market and pushing customers towards things that BlackBerry are good at them. And those three combined for me was – sort of the deciding factors.
Paul Treiber
You made a couple of points. You talked about foundational and BlackBerry uses, you know speaks with foundational layer. When you look at IoT, or specifically with an automotive, the software as it evolved over time, you know, BlackBerry person infotainment, you know, how do you see BlackBerry’s position? How has it evolved over time and where do you see it going?
Mattias Eriksson
That’s a great question and actually one of the biggest misunderstandings I get to answer when I talk to people that are not in this day-to-day. So if you look back five to seven years, BlackBerry QNX, which is the foundational core of BlackBerry IoT, used to be an integrated vertical stack in what we call the IVY domain.
So consumer experiences for sort of the digital cockpit, or five, seven years ago, the team way before my time here realized that that space was going to be eaten up by the digital ecosystem players. So Google, Apple, a lot of open source, and Linux and so forth. So the team, already five to seven years ago, pivoted the R&D from a vertically integrated stack to what we today call foundational software.
And what that really means is, you have a foundational layer of software that sits close to the chipset in each of the new compute domains and the evolving architecture of the car. And that’s where we are now having all the design wins. That’s where we’re getting traction. So think of ourselves as a horizontal place close – to the chipset for all compute domains for advanced IoT endpoints.
Paul Treiber
And where will we see in new cars being released, you always see the IVY system that may be profiled, it’s, you know, Android or Apple CarPlay. Can you explain like, when, when those systems, what’s the software that they run underneath? You mentioned the foundational layer is that always QNX? Are there other like well Android go that deep or like, well, Apple go that deep? Can you explain that ecosystem?
Mattias Eriksson
Another excellent question, so if you think about and this is playing out over multiple years, in different stages with different OEMs, but essentially, as you move from a decentralized ECU architecture to compute domains, and at some point, high performance compute what you get is centralized brains in the car managing certain domains. So and in these domains, digital cockpit is ADAS, as is another domain, body and chassis is emerging as one.
You know, power train in particular on the EV side is becoming a sophisticated software stack. The Gateway is evolving with multiple layers and so forth. So back to your questions if you hone in on one of those just take digital cockpit, which is where IVY evolved into in that digital cockpit very often, there is an Android stack or a Linux stack. And that Android stack, has certain things that are – what we call mission critical certain things that cannot fail.
Think of the digital cluster, for example, you don’t you cannot see a situation where your cluster with the speedometer and all the other information goes down and has to reboot. So under Android, there is and hypervisor and that hypervisor is typically QNX that’s our autos platform. And then depending on the implementation from the OEM, you might get additional layers of software on top of that hypervisor.
In the term – in the example of Android, very often, we also deploy what we call our advanced virtualization framework, which basically allows Android regardless of hardware underneath to access peripheral in the most efficient way.
So typically, underneath Android, if you take the new corporate domain, we sit with the QNX layer, the hypervisor, that hypervisor layer typically can have a couple of middleware components, virtualization, that could be a software defined acoustics platform. And that’s what we mean when we say multiple sockets, multiple layers in the evolution of the architecture protocol.
Paul Treiber
One of the starts of BlackBerry likes to mention is the EV BlackBerry had a status 24 or 25. Vehicles, what is it about EV that is leading to higher penetration of software? And is it EV in and of itself or is there a bigger picture trend going on?
Mattias Eriksson
Again, another great question, so the drive train is actually not that critical for the software architecture of the car. So regardless of whether drive train is – so whether it’s EV, there are multiple other compute domains that still require a sophisticated software stack. What is happening, though, and let me diverge a little bit on the tangent here. So five, six, seven years ago, we talked a lot about the case trends.
So the ACES trends depending on which acronym you like, so connected, autonomous, shared and electrified. If I look at the situation today, there’s one that is dominant. And there’s a second one that wasn’t on that list. And the one that is dominant today is electrification. And when I say dominant, I mean money being poured into building that capability from OEMs, and the ecosystem.
So you have electrification, which is very, very significant. And then you have the software defined car. And the software defined car sort of cuts through the other three a little bit. But it’s actually more important than that. All OEMs are today at different stages in maturity on a journey towards, you know, 10 years from now fully software defined car and that is a major undertaking, it is a massive, massive problem to solve for them.
So back to your question around electrification, so we do sit in the vast majority of electrified vehicle platforms, but we also sit in the other domains in that car. So the software stack for the drive train being electric is just one of the compute domains, the electrified cars still needs an ADAS compute domain, it still needs a digital cockpit.
And it still needs a body and chassis and so forth. It’s just when an OEM makes the transition to an EV. They typically try to fix everything else. So that’s why there’s a combination there. Sitting with EVs also, typically, it’s a little bit linked to the software defined vehicle.
Paul Treiber
Can we speak about the economics? In the past, as we see numbers thrown around, couple of dollars per vehicle? How do you think about the economics’ opportunity for BlackBerry with all these different domains, domain controllers?
Mattias Eriksson
I obviously think it’s fantastic. And the reason I think it’s fantastic is because you have, again, taking the history into account, you’re going from one fairly simplistic IVY stack, where we got multiple layers to, you know, five compute domains, each requiring an instance of QNX. And then for each of these stacks, you might need a software layer. And by the way, if you have problems developing software for this immensely complex endpoint, you typically want [Cylance].
You want to make it easy for your developers to develop applications higher up in the stack. So consolidating on one foundational layer makes it easier it reduces friction for the developers and that’s another Cylance for us. So let me stop there.
Paul Treiber
Shifting to design wins and I think there’s a lot embedded in that question. BlackBerry has a steady cadence of announcements around design wins, you know, some of the last ones in the last year or so Volkswagen, BMW, Volvo. How are these design wins different than perhaps what you signed in the past. You know, in other words, like, are they simply renewals for IVY or is there implicitly or maybe explicitly a larger, more strategic partnership there?
Mattias Eriksson
Yes so again, very, very important question. So, if I look at this business for the last three, we talk a little bit about this in the Analyst Summit in May. So what you have had the majority of our business used to sit in that IVY stacks for the last several years, you’ve had a decline of that business. And then you had this new business of foundational software picking up and obviously that decline is a drag on our business.
Our business is a long life cycle business, the royalties lasts for you know, 10 plus years and so forth. So as those in the IVY stack decline, you know, the others have to pick up both to cover that and to drive the next generation growth. So, so this notion of how do these design wins compare to the previous one. First of all, they are all now for foundational software, we do not win any – and we have no intention of winning any integrated IVY stacks anymore.
We win foundational software across the compute domains. And typically what happens is that they don’t just get – they take one domain from us, they take multiple domains from us. So the trends that we have been talking about for the last several years, they are now playing out. And there’s a significant different structural difference in what the previous wins were, and what these new ones are.
Paul Treiber
There’s always been a little bit debate, because there’s one automaker out there I’m figuring like to build software themselves. I think some automakers have tried to do that, you know, what have you seen generally a shift away from building it all and using partners? Like, what’s your perspective on that?
Mattias Eriksson
Yes, so there is some complexity yes, but the general trends would be the following. So if you asked an OEM five years ago, what they were going to do with this, they were going to build it all themselves. They were going to go open source, and they’re going to build the entire stack themselves for each of the compute domains. The reality is that even if you could, and let’s be clear, it’s a very significant challenge.
You’re talking hundreds of millions of code in the future of the software defined car, even if you could, is that really the best use of your resources? So if you’re an OEM and you’re trying to create a unique differentiated experience, based on that software architecture, you want to have your software engineers that by the way, you don’t have you’re trying to hire them, you know, thousands of them every year, you want them to be focused higher up on this in the stack on stuff that matters to the consumer.
So you want them to develop experiences and have login for the digital setup of that ecosystem that a consumer can see, in every release? Why are you spending time on the plumbing at the bottom of stack that allows essentially a very important function I mean, we provide decoupling of software and hardware. So very critical from a strategic point of view, but why are you spending time on that?
When there’s a solution that we believe is excellent, and that you just take off the shelf and by the way, it’s aligned with future roadmaps. So the general trend, from my perspective, and the big breakthrough for us, I don’t think it’s any secret was when one of the biggest proponents of automotive grade Linux, BMW, basically backtracked last year and said, look, let just stop. Let’s do QNX. And let’s focus on the experiences. So we see a trend where the OEMs are focused on the higher levels of the stack, and they lead the plumbing to a player like us.
Paul Treiber
We spent, I think, all the time, up to now it is focused around QNX, just wanted to shift to BlackBerry IVY. Can you just provide an overview of the IVY and what do you see as the opportunity for IVY?
Mattias Eriksson
Wow, yes. So obviously a great opportunity for us what is IVY? So simplistically and this actually requires a little bit of a discussion. But simplistically IVY is an edge cloud software platform that allows for dynamic deployment of sensor data processing at the edge, in particular, things like ML.
So think of it as a software platform, in that stack across compute domains. That makes it easy for an OEM or the OEMs partner to deploy ML models to the edge to process data at the edge and derive insights that they can use directly in the application the car or send back the cloud for whatever they want to do.
Big project for us joint development with AWS, one of the few large joint development projects that that AWS has, it is in the early stages, we announced it not too long ago making great progress on this development. And I think it’s a fantastic opportunity, with – lots of very difficult to replicate IVY, from my perspective.
Paul Treiber
Imagine similar QNX there’s this debate with the automakers? Do you build it yourself or do you use a partner? Is it a similar value proposition?
Mattias Eriksson
It is it is similar, although it’s slightly higher in the stack, so IVY doesn’t touch the chipset. So it’s slightly less engineering, geeky. It is still a software platform. And what it really does is it allows you to again, deploy that edge sensor processing in an easy way. So it’s about reducing developer friction.
And what we want to allow the OEMs and their partners to do is to spend less time again on plumbing, and more time on developing, you know, the algorithms that drive insights or consumer experiences. And don’t have to worry about you know, which hardware configuration it is and what stack you have and all that stuff. And by the way, in order to develop something like it make no mistake.
Forgive me if I’m going too deep – and now but many of the OEMs for some compute domains have these frameworks in place, in particular for digital cockpit. They spent a lot of time developing sensor processing frameworks at the edge that’s for one domain focused on one particular aspect of the future of the software defined car.
Doing that across the board in a future proof way is a totally different kind of game. And we believe that by helping them with this, we you know, obviously, reduce time to market, reduce complexity, reduce costs for them, and just make it easier and allow them to focus on the consumer experience.
Paul Treiber
Now, remind us again, I think BlackBerry mentioned that they expect design wins for IVY in 2023, I believe?
Mattias Eriksson
Yes.
Paul Treiber
How should we think about like the potential revenue opportunity, the revenue model, is it?
Mattias Eriksson
There’s no revenue this year, there’s no revenue next year for IVY, just to be clear.
Paul Treiber
Should we think about for the timing, the timing is it?
Mattias Eriksson
Yes, so the way the way we’re still chasing these design wins this year, we have had some challenges with COVID. So I think some of you might be aware, the first one, the first POC that was deep for us in the car Ford model was in China. And I don’t need to tell you the situation in China it’s been very difficult with the COVID lockdown. We are still saying we’re going to have design wins this year.
Let’s see – say yes if we’re there, I think that will be the case. In terms of revenue, the way you need to think about it is, again its embedded software. So similar to QNX, what happens is, first we sell the development framework and the toolkits, and then an OEM takes that and they start the development of their platform. And then they have to deploy that into a car for a start of production.
So this is a multiyear thing before it’s running in a car and actually generates revenue for them and for us And for those who attended the analyst Summit, you remember, we gave guidance for the revenue growth, organic revenue growth for this business unit. We did not include IVY in that three-year forecast just because it takes time to deploy and the initial deployments will not be large scale, as it’s a multiyear thing for us.
Paul Treiber
Shifting to macro in this year, when you look at like the IoT segments, such a pretty good growth through this year, anything greater than 20% or so what explains that growth despite the headwinds around, you know, supply constraints or automotive production? Like what explains that that growth that you’re seeing in the business?
Mattias Eriksson
Yes, so as usually in the real world, it’s a combination of factors, I think, fundamentally, the most important one is this notion that if you take, you know the full volume of cars and in that car volume, global deployments, of cars in the world, even if that one is flat or slightly declining, underneath, you have this evolution of the software architecture. So the software defined car and more software in the car is growing very rapidly underneath that total volume.
So the segments that we are in, you know, high end compute domain controllers, you know, ADAS, next generation digital cockpit, next generation gateways, et cetera, they are growing very rapidly in that overall bucket, and we are not in all cars yet. So, there is significant growth structurally underneath that overall volume. And then I think there are other bits and pieces, I think we – this notion of creating a separate BU focused on this has really helped with execution.
I think we have executed well, over the last 12 months, that one of the big things we did last year was we really doubled down on locking, including investments and timelines for a long-term roadmap, and that drove significant momentum in the assignments. What the OEM wants to do, they’re going to get into bed with you for 10 years, they want to make sure that you had a locked roadmap for the next five, and that they can rely on you being there and all of those stuff.
And that roadmap has resonated really well with the OEMs. I think that’s a big contributor to the design wins in the first half. And we typically announce the most important leading indicator only once a year. So the most important leading indicator for us is design win backlog that is basically royalties that are locked into contracts for the future. And in Q1, we announced $560 million royalties in the future.
I don’t think we necessarily have spoken broadly about the success in the first half. So let me do that here. In the first half of this year, we had more design wins in dollar terms than we have ever had for any previous full year. So in six months, we achieved more design wins than we have ever done for a full year historically so very significant momentum in spite to your point, the macro headwinds at the top level.
Paul Treiber
And I think you correct me if I’m wrong, but I think you see this in a near term in your revenues design wins through professional services and developer sees, how do you think about the timeframe from you know, converting those design wins into revenue?
Mattias Eriksson
So – that’s a great point. And we talked a little bit about it in the Analyst Summit. So essentially, it varies, but essentially you couldn’t use sort of a two year time of timeline from the sign win deficits [ph] and services to cause shipping and royalties, actually materializing. So, the design wins that we have now this year, they are typically for late 2024, early 2025 deployment that might seem like a crazy far away timeframe. But the benefit of that is you’re locking in very significant both revenue and profit for, you know, the next 10 years starting there.
Paul Treiber
And just remember like the supply chain constraints. I mean, that’s been a headwind to automakers. Have you seen – is there been any degree of them going backwards in terms of like, trying to move away from the amount of software deployed in the cloud, I was trying to push it out, because the supply environment?
Mattias Eriksson
Initially there was a little bit of de-scoping. So if I go back to last year, you could see, you saw some, in certain geographies for certain platforms, you saw a little bit of de-scoping for example, in the ADAS domain, because they couldn’t get what they wanted on the chipsets, et cetera. That obviously was bad for us. I think now, this is stepwise being resolved you’ve seen probably the chipset makers talking about how that is being resolved.
And it’s also sort of a mitigating factor to the overall macro conditions. So if the macro conditions in particular credit on the low end, making it difficult to borrow to buy a car, what you have on the positive side is, you know, the supply chain constraints for the chipsets are going away so you can deploy more of those advanced cars. And that is obviously good for us that that is maybe an additional points, that previous question you have.
Paul Treiber
Just want to open up to the audience if there’s any pressing questions for anyone, yes sir?
Unidentified Analyst
Just could be the end processing the key driver for that you guys see how far as vehicles and monitor separation of where in processing will meaningful done?
Mattias Eriksson
The short answer is no. So autonomous driving, I think, in general quite frankly, has pushed up. So if I look at, you know what the discussion was three, four years ago, and what the discussion is today, I think everybody’s agreeing, it’s slightly more difficult to put it mildly than what people expected. We do not need an autonomous driving stack to be deployed in order for us to get the ADAS domain.
Our ADAS the requirements for ADAS compute at the edge, you know, performance, safety, security and reliability applies as much to level two as it does to level three and level four, you know, not the driver, the driver is really the evolution of the software defined vehicle. And people really doubling down on making multiple compute domains, essentially software defined, and that’s what’s driving our roads.
Paul Treiber
We spent a lot of time on auto just want to just touch on general embedded market. What do you see as the synergies between auto and GEM? And what do you see as the magnitude of the opportunity with the GEM?
Mattias Eriksson
Yes, so I actually look at GEM as the sort of hidden GEM in our overall portfolio. So GEM traditionally was dominated by really low end ECU so and when you have low ECUs and MCUs being deployed. Typically, you end up with very basic, rudimentary software. What is happening in certain segments, in particular top end of certain segments is that the sophistication of the endpoint is going up very rapidly.
So more compute more sensors, more connectivity, more cloud interaction at the edge. And what does that mean? Well, it means that the requirements are similar to automotive. And this was one of the main reasons we were focused on automotive, we looked at automotive many years ago, we said, we believe automotive is the most sophisticated edge IoT endpoint in the world, the requirements are going to come there first.
When it comes to performance, no one has more compute at the edge than a car when it comes to you know, connectivity and sensors, all of it is leading in the car domain. So if we double down on fulfilling the requirements for the car, the others will come and that is literally what is happening. So if I look at, you know high end medical OEMs if I look at high end industrial automation, if I look at robotics.
They are all having very similar requirements put on the table and to give you a concrete anecdote, went back to Asia for the first time since COVID, the last couple of weeks, and industrial automation player in Japan showing the roadmap for us for the first time in detail for the last two years 2020 higher compute car requirements at the edge, then what we see from the OEMs so very rapid movement in the direction that we were hoping for. And I think there’s a great opportunity for us over the next several years in GEM.
Paul Treiber
You mentioned one of the changes that you’ve benefited from the moving IoT into its own business unit. What else have you done from like, a go to market perspective? And how has that helped design win momentum?
Mattias Eriksson
Yes, so a number of things. And obviously, the way it always plays out when you set up a new standalone business is that there are many things that you need to do, we need to pace yourself a little bit, if I take the go to market dimension, we have a multiyear program that is essentially about better reach. And the reality is that in the industries that we serve on a global basis, we are a fairly small player, we are a niche player, very focused.
We fit as an ingredient component into other players stack, it’s very difficult for us to cover all the various potential opportunities globally. You know, just think about how do you cover an OEM with all the compute domains and reorganizations that are going on. So from my perspective, we have way too few people in go to market, that doesn’t mean that I can change that immediately.
Because we have trade-offs. And so we are ramping go to market in terms of B2B sales, fairly significant data, 20% increase this year, probably going to be a 20%, 25% increase next year that’s one dimension. Another dimension on the go to market side that plays out in a multiyear perspective is the marketing. So think about BlackBerry overall, the marketing required to reach the enterprise software cybersecurity customers is very, very different from the marketing that is needed to reach the OEMs for particular foundational software components.
And just launched our first targeted, almost Account Based Marketing Campaign for acoustics. For those of you who are deep into automotive and are interested, that is an interesting product. It’s an interesting approach, acoustics sit in many different areas of the car, you know, not just in the cabin, but you know, for many other things that many acoustics components in a modern car, typically, they are spread out in an OEM across, you know, 510 departments.
So 15 OEMs, globally, let’s say that 10 people that really matters for acoustics in each OEM, how do, you reach 300 people globally, with a message at the right point, not an easy marketing task, and not an easy marketing task, the acoustics campaign we just launch is trying to do precisely that. And that is a, you know, one part of that shift in marketing approach. And then the third one so again, go to market expansion, direct headcount shift in how we deploy B2B marketing.
The last one is really ramping the partner support. So we work very closely with multiple value chain players, you know, chipset vendors, cloud players, technology vendors, that organization internally for us is too small. So I’m stepwise ramping that IVY is a great example of how we have increased that focus. So doing a multiyear R&D project with a leading cloud vendor, there will be more of that. But step-by-step.
Paul Treiber
We have one minute left, so we’ll take one last question.
Unidentified Analyst
QNX, IoT is 200 million run rate how much of that QNX?
Mattias Eriksson
So the vast majority of the IoT business unit revenue is QNX. And we still don’t break it out financially you will have to go back to Steve and John to discuss yes.
Unidentified Analyst
[indiscernible]?
Mattias Eriksson
So what is very important about that, and I – forgive my ignorance, if you go back a decade, I actually don’t know. But if I look back five years, which is how far back my research stretched when I did the research, what has happened and what I don’t think people fully realize is. What you asked initially here. So five years ago, 90% of our business was IVY integrated stack, one compute domain for automotive flowing through tier ones.
Today, that business is a rounding error. So although the business has, you know, maybe over a decade, I would say with five years not grown that much, there’s a massive structural change in what the business actually does over the last five years. So and over the last three, four years, that decline in our integrated stack has been a significant drag on, you know, the top line growth for this business. It’s a very different business today very different business today than it was five years ago.
Unidentified Analyst
Okay thank you.
Paul Treiber
I think we’re right out of time. So I want to thank both of you thank BlackBerry for being here.