Why AMD is staying out of the Internet of Things chip market (updated)

[Update: AMD announced John Byrne has left the company to pursue other interests — 1/12/15]

John Byrne has a tough job. As senior vice president and general manager of the computing and graphics business group at Advanced Micro Devices (AMD), he has to go head-to-head against Intel, the world’s largest chip maker. That means AMD has to be more focused and selective when it comes to designing its computer processors and graphics chips.

AMD isn’t doing much in the Internet of Things (IoT), which is Intel’s major expansion area. But Byrne notes that the PC chip market is still a $40-billion-a-year opportunity. He said that AMD has to execute on its upcoming Carrizo family of accelerated processing units (APUs), which will be focused on the mobile-computing market. About 300 million PC processors and 90 million graphics chips are sold each year, and Byrne wants AMD to get its fair share of those sales.

“I’m not in the business to lose money,” Byrne said, in reference to whether AMD will go into the Internet of Things chip market.

But Byrne does believe that virtual reality is just the beginning of a whole new wave of demand for computing products in the future.

We caught up with Byrne this week at the 2015 International CES, the big tech trade show in Las Vegas. Here’s an edited transcript of our interview with Byrne.

Above: AMD’s Eyefinity powers three displays.

Image Credit: Dean Takahashi

VentureBeat: How are things here at the show for you?

John Byrne: It’s great. It’s a bit crazy busy. I got here this morning, and you’re the ninth person I’ve met since then. But it’s always good to be in Vegas. I remember when I first used to come here. I had chances to tour the show constantly. Nowadays, that happens less and less. People just want to meet and meet, which is a good thing. This show and Computex are probably the two critical shows for me.

VB: What are you running at AMD?

Byrne: I’m an SVP and general manager of the client and graphics business unit. I’ve been in this role for five months now. Prior to that, I was AMD’s chief sales officer, and before that, I ran the global accounts — HP, Dell, Apple, Lenovo. Prior to that, I ran the channel and then the graphics sales and marketing team. My responsibility now is for sales, marketing, and platform engineering to bring all of our client and graphics parts to the market.

VB: How many years has that been at AMD?

Byrne: As I was going through that, I realize it’s been eight years, as of January 3. I had my own companies before. As you can tell, I’m not from Austin. I was based in London. I was fortunate that AMD acquired my company, and so I joined in 2007. That’s frightening.

VB: What are some highs and lows during that time?

Byrne: There have been lots of highs. I know you follow AMD and write a lot about AMD. There are some wonderfully talented people at AMD. It’s inspiring to wake up every day and be working with smart people. Not long after I joined, the acquisition of ATI happened. There was a lot of excitement in the marketplace. I remember when we brought Orochi [processor] to the desktop. But for me, I think when we started to really talk about our APU and bringing Brazos to the market and the vision of our APU. We were gaining share in the value space super fast.

There’s the work that we did resetting the company, with the three-step approach. The acceleration a lot of people talk about. The game console wins were significant highs. I think the Hawaii [graphics chip] launch was super high. I don’t think anyone anticipated how well Hawaii products were going to perform in the marketplace. Those are some of the highs that I’ve enjoyed.

Our execution on our server roadmap was a bit of a low. Llano could have been better. But AMD is a challenging, but inspiring, place to work.

VB: Going forward, what are some of the more focused areas you guys are excited about?

Byrne: I don’t know how well you know the restructure we had. We created these two uber-business units, client and graphics. There’s what I call the classic, traditional AMD, and then you have the ESE. When I took over this business group, they were two individual business groups. They had talented leaders, but they were individual in thinking on the roadmap, individual in thinking on our ISV [independent software vendor] strategy, our benchmarking strategy, our go-to-market strategy.

One of the reasons we pulled these two business together is we believe there’s significant synergy between their products. If I believe the future of [computing] is gesture and touch, and then VR and AR and immersive reality, it leans heavily into a graphics and client roadmap. I collapsed the businesses into one, very quickly. I set up a new product leader. That product leader now has someone who’s focused on client and someone who’s focused on graphics. We’re driving to clear accountability into the marketplace. I centralized our product marketing, centralized the go-to-market, so we had a consistent message. I also centralized a lot of our sales function and our benchmarking team. I centralized a lot of our ISV team.

We spoke a lot about [AMD’s application programming interface] Mantle. Mantle is important, but it’s ultimately what experience you want your customers to enjoy. Mantle is allowing you to execute to that and works with the relevant games … to get there. Where we could do better is ensuring that we’re enabling the relevant experiences and solutions. The hardware is critical, but do you have the software, the ISVs? Do you have the platform off in the market? Centralizing my ISV engagement has been critical.

Above: AMD-powered gaming computers at CES.

Image Credit: Dean Takahashi

VB: I look at Intel and their shift into things like IoT and wearables and the small Quark chips. They seem to be doing so much more of the work that their customers used to do. I wonder how you guys can either adapt to that or learn some of the same things but on a smaller scale or in a more focused way.

Byrne: They have jumped to Quark and IoT, but even if I look at the classic PC market for a minute, there’s still 300 million PCs, still 90 million graphics chips. If I look at Intel, Nvidia, and my revenue, that’s still a $40 billion market — even before you get to the IoT. If you look at the gross margin profile of that business, it’s still significantly more than AMD as a company’s average. There’s still significant market opportunities in the classic PC space.

There are some things we have to do well there. We have to continue to work on our x86 performance, ensuring that each product we bring to market is better x86. We have to maintain our graphics leadership. For notebook, we have to ensure that we’re driving to improve battery life. We’ve done well, but I think if you look at Carrizo, it’s a quantum leap for us in terms of power efficiency. Those are table stakes. It allows us to play in the segment — attacking i3, attacking i5. Again, i5 is a quarter of the market.

I believe that we have not historically played in commercial. We’ve won the industry’s largest single tender in commercial 18 months ago in India. We won Elitebook … with HP last year. Wait until you see the lineup of commercial platforms I have with Carrizo. It allows us to continue to attack that i3, attack that i5 consumer, and really get to penetrate the commercial market space. We’ll attack graphics. That’s going to be my strategy next.

Of course IoT is important. The question that you and others are asking is, hey, if you have a wearable, there are two ways to make money from it. One is the silicon inside the wearable. That’s under $10. You’re seeing that with Quark and some of the other investments our competitors are making. I’m not in business to lose money. Share and revenue is nice but so is profitability.

All of those devices have to be connected. I believe that we can have an answer on IoT, be it through data center or storage, to enable IoT. That’s what I see us going after. The low-cost pieces of silicon in IoT our competitors seem to be doing are not part of our strategy.

VB: What are some priorities? There are things like the ARM-based server chips in the works. There are video game chips. There’s the traditional road map as well. Where would you say the resources are going?

Byrne: When I look at the world of virtual reality, what’s being done with [head-mounted displays] at Oculus and Sony and Samsung, it’s just the beginning. I really believe that. I think the future is immersive reality. It’s going to need significant [computing]. In the early days, it’s going to need significant graphics capability. It’s going to need super frames per second. It’s going to need low latency. The products we’re bringing to market will penetrate those.

That will evolve. It’ll evolve from a desktop with a big chunky graphics card, going into our APUs, and then into something else.

VB: I’ve heard people say that maybe you should get to 120 frames per second on virtual reality before you can do away with the lag and the seasickness.

Byrne: It’s got to be a minimum of 100. The last thing anyone wants is to put on that headset for the first time and feel sick. You’re not going to rush back to it. It’s smart to do things properly. I see us going after the immersive world. What I’m putting a lot of my resources into … there’s better x86 and better graphics, but the sweet spot of the market is attacking that core i5. That’s the biggest part of the PC market.

We’ve done very well. The A10 did very well. We’ve captured share against i5. But we’re still nowhere near enough. I spend a lot of time on software — getting the relevant ISVs, getting the experiences. You’ll see touch and gesture enabled. You’ll see content manager enabled.

Another thing you’ll see me spend a lot of time on in our graphics business. We have great graphics hardware. I talked about Hawaii. But one thing we should have improved on is the experience of using an AMD product. Our drivers were good, but they weren’t good enough. I applaud the software team for the recent Catalyst Omega driver, which has earned multiple millions of downloads already because the hardware is just part of the story.

Where I want to then go is … if there’s 300 million processors and 90 million graphics chips sold per year, I want our technologies to be talking to each other. If my APU and discrete GPU can talk to one another, that gives me a true advantage that our competitors can’t do. We call that A plus A. That’s where we’re spending a lot of our resources and talent.

Above: AMD-powered Advantech GMB-M1 gaming all-in-one desktop at CES 2015.

Image Credit: Dean Takahashi

VB: VR sounds like it drives the traditional part of the business first.

Byrne: That’s the way I look at it. Once you try the Crescent Bay, it’ll blow your mind. I tried the Oculus 1, and it was great. You thought about it as more of a game-playing device. When I saw Crescent Bay, though … my family lives in the U.K. When you see an environment where I could literally be seeing my parents, and we’d appear to be sitting down having a conversation in the same room. If I think how that could evolve into automotive or medical super fast, the VR we’re seeing today is just the beginning.

The very first iPod was 15 years ago now, which is frightening to think. Think about how far that technology has come along in 15 years. The VR, and ultimately the immersive reality, will take time, but it’s a place where we can play. We want to be leading in gaming. VR’s going to have significant impact on gaming. … Photorealistic VR is a one-million-times speed up from what the current GPU can do. … When you talk about how we drive high-performance graphics, that’s the goal, photorealistic VR. It requires a lot of horsepower and a lot of work on the software side, but it’ll happen.

VB: You had 700 layoffs or so. Did you cut any particular pieces of the larger effort in order to do that?

Byrne: It was my business. I purposefully took the biggest impact from what happened recently. If you’ve been at any other big companies, a reduction in force means everyone takes a haircut. We didn’t do that. As I mentioned at the beginning, I had these two separate businesses — two different product management teams, two different ISV teams, two different benchmarking teams, two different go-to-market teams. I had two of everything.

I thought about our customers, getting close to our customers and bringing products to market that customers really want. That’s what matters. When they win, we win. It’s critical to listen to the voice of the customer. But you have to be simple to do business with. You have to be fast, agile, nimble. People are held very accountable.

Looking at my business, client and graphics, I let go a third of the VPs involved in the business, a quarter of the directors. I took a whole chunk out of middle management, which is unusual. Normal that appears to be a protected species, but here it wasn’t. I wanted us to be faster. I no longer have two different product teams. I have one. People know where to go. When you’re the underdog, you have to have that underdog mentality. That’s why I did it.

We protected the road map. Nothing came off the road map, which is important. We protected R&D. I wanted us spending more, especially on software. Our hardware has always been good. I want to continue to work on our software optimization. There’s still hiring going on in key technical areas and other parts of the business.

The CTO’s office has been driving a significant re-engineering of that function. We’ve been moving away, each year, from doing discrete, monolithic CPU designs to a more modular approach. It simplifies the design process and allows him to get more done with fewer resources. That’s one way we’re able to protect the road map and invest in the future with lower resources.

If you look at some of the things that have to happen, Carrizo has to be a great product. We’re feeling very excited about the performance that we’re seeing in Carrizo. It’s behind closed doors here right now. The battery life is looking great. The performance is looking solid. We’re pleased with the designs we have already with the top customers in the marketplace. The ability to have Carrizo as one [system on a chip], one infrastructure, scaling up and down the platforms for our customers is critical. You see all the big guys here. You know how much it costs to get a platform design developed. And we’ll be bringing a king-of-the-hill graphics product to market. We’re excited about it.

Above: Tiny computer powered by AMD.

Image Credit: Dean Takahashi

VB: How is it working under (AMD CEO) Lisa Su compared to (former CEO) Rory Read?

Byrne: Lisa and I have worked together as peers for three years. She became my boss for three or four months as COO. Lisa and I have a great relationship. If I was to articulate the difference between then and now, Rory is an operational wizard. Being the COO of the world’s largest PC manufacturer at Lenovo and getting an operation moving super fast and then changing directions super fast — Rory was brilliant.

The level of discipline that Rory brought to the company … by discipline, I don’t just mean the balance sheet and profitability, although those are big things. But even assuring that we brought the products to market. Because he came from a customer. AMD had these great products, and then they’d have a slip. A great product, then a slip. It impacts your ability to get share. Rory brought in that level of execution.

With Lisa, Lisa obviously has operational discipline, but being a technologist and an engineer, she has a true passion for the technology and the products we’re bringing to the marketplace. That’s a subtle difference. Rory got us in a great fitting. Stage one of the reset, acceleration, getting us into the beginning of transformations. Transformations are never linear. I wish they were, but they’re not. Lisa is the right leader at the helm for AMD.