raj talluri

Video game makers haven’t always been thrilled about making mobile games. But Qualcomm hopes to change that. Raj Talluri — the senior vice president of product development at the San Diego, Calif.-based mobile chip maker — addressed the DICE Summit, an elite game industry event in Las Vegas, to wake them all up. Qualcomm’s Snapdragon processors have been used in more than 500 million Android devices. And smartphones are selling at a rate of more than 500 million a year. That’s a pretty large market to target.

Soon enough, it won’t just be numbers that are attractive. Mobile chips are coming soon with as many as eight cores. That means that these low-power chips will be able to come close to the processing power of current-generation consoles on smartphones, tablets, and smart TVs. And mobile networks will be carrying data at 150 megabits a second. That will make your cable modem look slow.

We interviewed Talluri at the DICE Summit. Here’s an edited transcript.

VentureBeat: So your talk was telling everybody about the size of the opportunity out here.

Raj Talluri: Yeah, and the technology that we do.

VB: Do you get any debate about that right now? In years past, these guys would have looked at the smartphone as a platform that couldn’t do what you wanted it to do. They would have ignored it. There are people who experimented with smartphones, and they did very well, but for the most part, the more hardcore teams were still working on consoles. What’s interesting about the timing right now, and how this is changing? Are you seeing some of the game developers change their minds?

Talluri: It’s a couple of things. One is performance, what processors have done in the last few years. They were like little toys before. You couldn’t do anything compared to what you can do now. That’s one thing. The other aspect of it is that the market opportunity is much bigger now. Selling 500 million smartphones. Another thing we’re finding is that some of the traditionals may not be doing it, but almost every one of them at least has a division that does mobile. They’re getting into it. And there are newcomers that have made a killing.

Obviously, there’s money to be made. The numbers are real. For a long time, it was only on iOS, but I think the Android Marketplace is working out. I talk to people now, and they’re able to make money on Android. Discovery is working. Operator billing is working. Free-to-play is working. Lots of things are changing now. But we’re still at the beginning. There’s a lot more to come.

raj 2VB: Does the ecosystem seem fairly complete to you?

Talluri: No. The ecosystems can always be better. I think it’s still in the beginning stages. Honestly, to me, this concept that you have to buy a game or get a game, and it comes on one platform one day, and then you have to buy it on another platform — in time, it’ll go away. I mean, how many times have I bought a game on Xbox or whatever, and then I have Windows Live on my phone, and it doesn’t work together? What happened? Why don’t I get the mobile version when I get the other version? Those kinds of things should go away. There shouldn’t be any real reason for it. I think all these things will happen with time.

VB: I find I’m looking forward to 150 megabits a second than anything else. [Laughs]

Talluri: It’s on the way. We showed it in the booth. We had it running.

VB: What is that going to do for gaming on mobile devices?

Talluri: Clearly, latency, when you play — it’s not just connected to bandwidth. It’s also about how the network is structured and how much data you can transfer. Those are the two big things. I think that this whole client-powered thing will become bigger, too. We’ll perform a lot of the work on the phone. We’ll do some more of it in the cloud and then ship it back and forth. It’s no longer about having all the processing in one place. That’s a paradigm shift that will happen.

VB: Something like OnLive, where you can do more cloud gaming?

Talluri: Yeah, but not just cloud gaming. Some of those things are not very scalable. You put so much computing in one place, but you can’t scale it. You undertake it when you have a powerful computer on both sides, and high bandwidth in between. It’s distributed gaming, more than cloud gaming. That’s how I think about it. You have a pretty powerful processor on this side, too.

VB: Predicting when we’ll get 150 megabits per second of bandwidth is a hard thing, though, right? You build the capability into the chipsets, but what’s AT&T’s schedule like, or Verizon’s schedule? Once they build it, will the capacity be upheld? The capacity of 3G and 4G seemed great until people started using it and sending YouTube videos all over the place, and then it crashed the whole system.

Talluri: Well, I don’t know if it crashed the whole system. My 4G still runs pretty well. [Laughs]

VB: It stressed the systems in some ways, though. I can’t get good voice service in a lot of places now. I would point the finger at all the data going everywhere. Anyway, is it hard to predict when this new bandwidth era is going to happen?

Talluri: Not really. It’s fairly simple. You look at LTE, where we were having the same conversation. We’re going to drive 40 megabits a second. I’d see the same question from Engadget or somebody about when we’re going to get it, how we’re going to get it, where we’re going to get it. I mean, right now, in the U.S., you can’t sell a smartphone if it’s not LTE. It’s been two or three short years from time we set out LTE. Now it’s ubiquitous. It’s moved very fast. So these deployments happen rapidly. Operators want to grow their business. They want to invest more.

VB: Verizon is another question, really.

Talluri: Yeah. It’s true that all those things are there. But the reason that you’re not getting it is because people are using it. That means they’re fulfilling a need that exists. Your tolerance level, for example — like when I’m on LTE and this guy sends me a 20-meg Powerpoint and boom, I’ve got it. Nobody even thinks twice about shipping a 20-meg Powerpoint, knowing that I’m on a phone and I’m the road. Nobody worries about it because they know I’ll get it. They know I’ll see it on my smartphone. I’m checking the Powerpoint right back there before the talk. It’s pretty nice.

My point is, are you getting the promised 40 megabits per second for everyone we said we would? No. Are you getting a ton more than you got before? Yes. Have you changed the way you operate because of that? Yes. The same thing is going to happen with this. There’s not much point in fixating on this 150 megabits per second number. Think about four times more than what you get today. Your speed will be four times better. Then, when you eat that up, we’ll have to do one more. [Laughs] That’s how we make a living. We create the need, and we fulfill the need.

raj 3VB: All of these different features you talked about — how would you rate them as far as what’s going to be most important for game developers and gamers on mobile devices?

Talluri: Right now, I think that game developers on mobile devices are not really taking advantage of all the capability we’ve put in there — all the sensor features, the cameras, the UI, the gestures. Most of them, they just take their console game or some other game and port it to use a touchpad and a small screen. That’s why I liked that “multiplatformization” talk. It’s the first time I’ve seen people realize, “You know what? I don’t really have to just do that. I can do it in different ways.”

People want to hold a tablet in one hand and play with one hand. I think you can do a lot of things like that. The GPS in here is so integrated. You can use locations in so many interesting ways. There are so many sensors in here, and people don’t use any of those. People could use voice to control a game. They could use the gyros to control a game, or the motion and altitude sensors. That’ll happen as more and more people realize that it’s a viable platform for making money, and as awareness of all the things we’ve packed in there grows. We’re doing our part by providing APIs and so forth so people can access those features. Better and better games will come along. We’ll get to the point where a lot of things will be developed for mobile first, instead of ported. It’s just a question of time. We’re pretty bullish.

VB: The implications of things like 150 megabits per second — you could get rid of your home Internet, right?

Talluri: Yeah.

VB: Who needs it once you have mobile that fast? How much change are we going to see on that front, as far as mobile breaking out of whatever ghetto it’s in right now?

Talluri: [Laughs] It’s a pretty comfortable ghetto, I have to say. The way I see this is that a lot of people look for something to replace something else. It seems to be the natural tendency as far as how people look at things. “Will mobile make my Internet go away?” The way I like to think about it is … as a user, when I’m at home, I don’t want to think about how I’m connected to the Internet. Am I connected over mobile? Am I connected over Wi-Fi? Am I connected over Bluetooth? Am I connected over a cable modem? Why should I care? Somebody should just provide me a service. I come home and it works.

It already happens like that in the home now. You don’t think about whether your phone’s on Wi-Fi. You turn on your Wi-Fi, set your access point, come home, and it works. I think that’s how it’s going to be. When you have a better connection like LTE Advanced or whatever, when it’s got a better economic model, you’ll get that. If you’ve got a cable modem that’s working, you’ll get that. Or you’ll combine the two in some way. I don’t think any one thing will replace the other. All these things will work to a point where the consumer benefits by getting good speeds.

VB: Do you feel like the consumer and game developers do have to get to know Qualcomm as a brand? I guess the model before you is Intel Inside … .

Talluri: We don’t worry too much about that. We don’t necessarily want them to get to know us for marketing’s sake. That’s interesting, but not important. What’s important is that since we make our own GPUs and make our own CPUs and design our own processors, we want developers to get to know us so that they can optimize their games to use everything we offer them, so that the experience for the end user is better.

The better the experience the end user has, the better the phones sell. The more chips we sell, the more games the game developers can sell if everything works out. That’s our goal, to make sure they know what they do and provide them access. That’s why one of my slides was “Access Is The Key.” It’s happening already. For example, if you go to Android Marketplace and you download a game onto any of the Snapdragon phones, they all run beautifully. Some of them, we never even talked to the developer. They just work.

VB: The capability that you’re providing here — are people taking advantage of it as fast as you would like? Does it happen naturally, or do you have to do a whole lot proactively to seed this community and make things happen? Like you could invest money into game developers or whatever. How much is already happening naturally?

Talluri: Quite a bit is happening naturally. There was a game, Need for Speed, that someone showed on their platform at CES. I asked my guys, “Hey, how does that run on our platform?” We downloaded it, and it works beautifully. We never even talked to the developer. What’s happening is that because we have such a large footprint, the developers are automatically making sure that their games run well on Snapdragon before they put them on the marketplace. If it doesn’t run on Snapdragon, it’s not going to run well on the majority of phones.

We have almost all of the high-end products, like Galaxy S3, Droid Nexus, Droid Maxx, or Nexus 4. Any developer working on Android usually does that. Anybody working on Windows automatically does that because we’re the only one offering Windows now. The new Blackberry uses us. Just by default, we’re getting a lot of that. But still, we like to make it easier, so that people can talk to us and work with us on our latest chipsets before the phones come out. When the phone comes out, they can take full advantage.

At CES, for example, we showed a couple of games from Gameloft — The Dark Knight and Asphalt 7. Those are two really high-end games. Both of them were running in full 1080p at 30 frames per second on our latest processor, taking advantage of quad-core. We worked with them to optimize that. Then we get all the optimizations back out, and now all of Gameloft’s games will use them. We also spend a lot of time with game engines, like Unity and Epic and so on. Once the engine is optimized, all the games on top of that will benefit.

VB: What does eight-core get you? If quad-core gets you 1080p …

Talluri: [Laughs] I don’t know if quad-core got us 1080p so much as the combination of the cores and the GPUs and the memory controllers and the video engines and the audio engines. It’s a whole system, the SOC. That’s how we look at cores. We launch them and we put in as many as needed to provide a real benefit. We haven’t talked about eight-core. We did say that the eight-core in Exynos was a little bit more marketing than real eight-core because you have four big ones and four small ones. They had to do that because the power on the four big ones was too high.

VB: Folks like Nvidia are taking interesting steps like creating their own game machine. One of these days, Intel may get so frustrated that they’ll make their own phones. What do you guys think of that?

Talluri: Competing with their customers would be a bad idea. Usually that doesn’t work out well. We try not to compete with our customers. We’re happy doing what we do, which is provide the processors to people who want to make phones.

VB: It’s interesting to see these new challenges that blend the console and the mobile device, like Ouya or Nvidia’s thing. I wonder how well this space is going to do. It’s bringing mobile to the TV.

Talluri: We’re seeing a lot of traction for people taking our Snapdragon processors and putting it inside the TV, or putting it in little boxes that sit next to the TV. There’ll be a whole bunch of those that you’ll see launching this year, kind of like the Ouya. A lot of people are doing that, but not just for games. They’ll play video and audio. They’ll run Android Marketplace and so on. It’s more like a smart TV or set-top box. Of course they’ll run games because a lot of the games on the marketplace are already optimized for our processor. But we’re looking at things like that rather than a dedicated gaming device.

VB: I had a look at some stats last week. They said that 95 percent of all content purchases on Android mobile devices in Korea are games. In the U.S., it’s 76 percent. I guess that’s why you guys care about this. [Laughs]

Talluri: You want to think about what people can do with your stuff, right? We care about what people use our processors to do.