MAY 21, 2012 • It is easy to loose track of all of the game related initiatives that have been born at Qualcomm during the last decade. The company has been prolific at leveraging its mobile chipset technology to encourage and facilitate mobile phone gaming. The downside of this wealth of activity is a perception that the overall approach is scattered rather than directed at achieving tangible goals.
We look at things differently, however. There has been so much revolutionary growth in mobile technology since 2000, that just to keep up requires rethinking major programs on a regular basis, as well as launching promising new technology at a breakneck pace.
To tie together the loose ends and get an up-to-date look at how Qualcomm is leveraging games on mobile platforms, DFC spoke with two executives at the Southern California company: Dave Durnil, director, advanced content and gaming; and Sean McCann, marketing manager at Qualcomm Atheros.
DFC: Going back over the last ten years Qualcomm has been very prolific in developing game related technology on mobile platforms. We’ve seen one initiative after another from BREW, UI ONE, Media Flow, Q3 dimension, QX Engine, Zeebo, Snapdragon, Game Command, Killer Technology. What we would like to know is what has been the guiding direction behind all of this activity, if there has been one, or if it is more along the lines of different teams coming up with great ideas and operating more independently.
Dave: I can give you the highlights. Some of the technologies that you outlined in the question are completely different and they might not necessarily be tied together. Media Flow, for instance, that was a whole separate business unit that spun off on everything else. The way that it has been a common theme has just really been the progression of the technologies as well as the market as it evolves.
Ten years ago we were actually one of the first pioneers in mobile gaming in terms of really helping the space evolve. This went back to the 2D games, this is on the flip phones, this is really helping try to enable the carriers in Japan, who initially were are some of the first
adopters of the ZMA networks, to also look at gaming and how we can help and even back then in the early 2000, 2001 and 2002. We invested in Jamdat, which is actually got acquired by EA and that helped form EA’s mobile division. We invested in a lot of other gaming companies in terms of content and other opportunities and partnerships just to spin up some of the mobile gaming units within these major publishers.
From the very beginning we saw a need to figure out these new channels and new technologies – on how to bring gaming to cell phones at the time and now to smartphones, tablets and also now to TV on through smart TVs. Through that progression there’s a lot of things that have happened at Qualcomm as well as in the market but one that happened in the very early days is that we did see a need of putting dedicated 3D hardware into the actual chipsets. We then spun up internal efforts to start building out 3D graphics hardware. Qualcomm is actually one of the first companies to integrate some dedicated 3D graphics hardware into their actual cell phone chips and get those shipped in the world. You can see part of that progression in things like Q3 Dimension – that was just a brand name very early on – just to refer to some initial integrated 3D hardware that we had in our chipset.
Once we acquired the API handheld group from AMB a few years ago and brought over some original graphics architects that worked on Xbox 360, we ended up with a lot of those guys who have the core commercial, high-end PC graphics card experience, as well as the handheld graphics experience plus the original architects that worked on Xbox 360. Having acquired all of that talent and technology, we were able to integrate our own dedicated GPUs fully into our processors. That’s a big leap there now obviously has to obviously be another change in the branding. This scenario happens over and over again as our strategy keeps evolving and the market keeps growing.
When we did the QX engine there wasn’t really at that time a nice dedicated, optimized game engine for pre-smartphone gaming. We weren’t going to go build a whole game engine but we wanted to be able to enable a lot of the gaming guys out there. QX engine was a really nice low level mesh animation engine and it was highly optimized and tuned for that particular hardware that we had at the time. We actually provided that source code we had sort of a tool chain that could integrate with existing tool chains that were in game studios that were out there. We never formally commercialized QX in terms of a product, but ended up just really getting it out there for a lot of the major studios who wanted to know how you would do this low level mesh animation optimization on our particular hardware in a lot of those simpler 3D games at the time.
We also wanted to make it really easy to take some existing content and create some type of a mobile version that the content owners could run on a cell phone and deploy. That’s how we developed BREW. On the gaming side, when BREW first came out, we had Java and we had BREW. BREW was really this huge opportunity for game developers to finally have native gaming and native gaming text capabilities on a cell phone before smartphones, as well as the means to distribute those games. BREW alone paid out something like over $3 billion to developers as part of that program. It was one of the first globally deployed app stores in the world. BREW actually provided that capability before other people did and that of course spawned new markets and encouraged other people to build their own app stores and their own businesses around that type of model.
UI was a different effort that was really to enable taking advantage of some additional 3D hardware as well as the full multimedia capability and getting the tools to device manufacturers so they could build their own custom UIs. At that time, way back then, OEMs differentiated by the devices they built and also by the look and feel of that device. We were trying to get a more high-end UI to get away from that standard kind of scroll rolodex menu that used to exist on cell phones.
Our general philosophy has been to keep evolving and making mobile a better platform and a better experience for everyone. Since day one we saw mobile gaming was going to be big, we wanted to be a part of it, and so we helped enable the whole market.
DFC: What is the concrete benefit to Qualcomm? Do mobile games really sell more Qualcomm chip sets?
Dave: The benefit is to really bring a bigger and better gaming experience to the consumer. We focus not just on the hardware but the ecosystem. We put all the high level OS’s that are out there today on our platform except for iOS. But there’s quite a few others: there is Android, there’s QX with Rem, there’s Windows Home 7 and Windows 8, there’s BATA, there’s different WINEX variants… it just kind of goes on and on. But the whole idea is to bring the best experience we can to consumers and make sure that that experience is optimized and tuned. If we can bring that to the consumer then that indirectly helps us sell more products.
So last year one of the big initiatives was to box the Snapdragon Game Pack. That whole concept was about how we could bring some bigger and better games over to Android for everyone to enjoy. Our Snapdragon processor platform has support for high-definition graphics built in. The Snapdragon game pack was initially intended to feature 100 optimized games for our Snapdragon chipset used on Android phones. We got a lot of feedback from consumers who were very interested in the Snapdragon game pack. They were asking, “Well out of all these games you have in your game pack, how do I know what games in that pack are optimized for my particular device?”
From that question is really where Snapdragon Game Command came from. Game Command became a nice discovery app where consumers could download it, and depending on their particular device, it would search for all the games in the game pack, and Game Command would return with all of the available games optimized for that particular device. Game Command provides an extra level of filtering for each device that consumers couldn’t get from Android Market.
There are a few other useful or meaningful features in Game Command such as group capabilities so you can group all of your games under one icon where they can be sorted. You can display the last time they were played, or by alphabetical order, or by publisher. We are going to keep evolving Game Command by adding new features that will hopefully be more useful or meaningful to the consumer. so far we’ve gotten a lot of good feedback back on Game Command and we are going to keep taking it forward.
DFC: When these initiatives for enhancing mobile gaming get pitched internally, what is Qualcomm’s prerequisites or standards green lighting a new technology project?
Dave: We don’t necessarily share our internal processes on how we approve certain things. A lot of it actually is driven by market trends and newer features that we think would be important to bring to mobile. For example, the PC platforms going mobile, whether they are coming to tablet or other devices and how we leverage the different types of technology around PC gaming. It’s just not having comparable quality or comparable types of experiences, but then how can we make that experience more compelling or maybe more unique on a smartphone or a tablet?
DFC: When you roll out something new- whether it be a BREW, Killer or Snapdragon – give us a feel for how you invest in partners, developers, and carriers. How do you ramp up support for these partners?
Dave: Well, we already have support in place because we’ve been working with all these major guys for a very long time. So there isn’t very much of a delta there in terms to ramp up – we’ve been doing that since day one so that’s already taken care of. I don’t want to say so much detail on exactly how we get those guys ramped up but it’s going to be different for everyone. We work with all the major game publishers out there. We’ve been working with EA, we’ve been working with Namco, Sega, and Konami – all the major ones – but then also a lot of the mobile ones like GameLoft and others. We’ve also been focused on a lot of indie developers.
There’s a lot of great major publishers out there that have interesting concepts that they want to bring over from other platforms. We obviously want to open up a channel for that, and bringing on the Sanpdragon Game Pack was a vehicle to provide that channel for a lot of them. As for the indie developers, a lot of them are trying to figure out how to get into mobile or how to bring their games over to our particular platform. These might be games that exist on Facebook or on a different device. We’ve been helping those guys bring a lot of that content over because I think there’s some really interesting, unique experiences that they can offer. And so supporting them is different than lets say with one of the big guys right.
We also focus on the middleware guys, the guys that are building the engines, because they do have a bigger impact on the ecosystems. We do have a dedicated gaming team here at Qualcomm both on the business and engineering side. The engineering team does focus a lot on middleware and then working with a lot of the key developers and optimizing and tuning. A lot of developers have their own engine and so they know they want to turn on additional features they want to take full advantage of their hardware. A lot of that information is available on our developer website but depending on who it is and what it is we also provide direct and dedicated gaming expertise for reporting and optimizing some of those over on our newer hardware when it comes to market.
DFC: What are device manufacturers telling Qualcomm they want in relation to delivering games on to their hardware?
Dave: I think in general they want what consumers want. They want the best experience. They want gaming anywhere, anytime. It’s really up to us to figure out how to deliver that experience in a product they implement and would be valuable to the consumer.
DFC: And carriers, what are there concerns? What are they asking for?
Dave: You know, carriers want pretty much the same thing but they also want some differentiation just like the customer. We look to see how we can help them and enable them for whatever particular product or market they are in. Carriers are interesting because they have their own app stores and they’re obviously trying to figure out how they get a piece of the pie. We do a lot with our technologies and through some of the gaming and content to enable them more.
DFC: Moving on to something completely different, the Zeebo console for emerging markets. The system was funded by Qualcomm and it used a significant amount of mobile technology that came from Qualcomm. And now, as far as I understand, there’s a new system being developed. How much oversight does Qualcomm actually exert over how Zeebo developed and is still developing?
Dave: Well, Zeebo actually was an idea that came out of the Qualcomm innovation network here internally and I was actually part of that effort. To realize that idea, we spun it out as a separate startup company. Zeebo is an incorporated company, it’s a completely separate entity and they control their own destiny. Qualcomm was an investor and we sit on the board, but we don’t influence or control the actual design or day-to-day business. All of that is done by Zeebo business management.
Zeebo in the very beginning really was to target this low cost, dedicated gaming console for emerging markets. Sort of the next billion gamers or consumers out there. If you look at a lot of the high-end consoles, the overhead for import tax was significant. You could be paying a thousand dollars to get a seventh generation console in those particular markets. So the general idea was to bring a low cost, okay gaming console that would circumvent the piracy that runs rampant in those regions, a console that could download all the content through a 3G network and secure it on a device without physical media yet making that content secure. We were thinking in terms of bringing that type of experience, or at least making that available, to a lot of people in those regions who might have one cell phone they share for the whole family. That’s a whole different market when you talk about the emerging markets in terms of experience or what would be good enough for them. And so Zeebo was really designed to be a good enough console for particular regions.
DFC: We understand the next Zeebo is going to be Android based platform. How involved is Qualcomm in terms of helping with that technology development?
Dave: Well, to the degree that we have to be involved. I mean we provide the platform as well as ports of Android to our platform so they’re just like any other customer that would pickup our particular chipset or build. Beyond that we don’t do anything extra.
DFC: Your Killer Networking hardware. What was the original intention for Killer and how is it developing? Were you really looking for a way to drop latency games or were you more interested in enhancing video streaming and such on mobile platforms? Give me a feeling for what Killer was supposed to be and how it is actually turning out.
Sean: Killer technology has existed separately for a few years before Qualcomm purchased Atheros in May of 2011 to form Qualcomm Atheros. Then Qualcomm Atheros bought the company that I worked for, BigFoot Networks, in September of 2011. BigFoot had been in the market for about five years prior to that making PCIe Ethernet cards optimized for low latency gaming and a variety of other. Gaming primarily because it’s easily the most popular, highly installed and latency sensitive sort of traffic. But eventually the intelligence that we were developing to run on the card and around the card allowed us to control and manage the latency in pretty much any application that ran on a Windows PC. It was this software and these sort of streams of intelligence that we developed that allowed us to develop better and much better performing wireless hardware for laptops starting in February of 2011. We had PCIe standalone cards as well as implementations on gaming motherboards, the latter we are actually still shipping with one of our gaming partners, Gigabyte.
So the company was founded with an eye towards gaming because when it started in 2006 and 2007 that’s what the founders were interested in, that’s what the market place looked like at the time and there was a lot of latency built into the Windows network stack that treated all traffic equally. With many tweaks and some innovative technology we were able to get game traffic around the Windows network stack and back out to the internet faster than any other network hardware out there. So that’s how it started. By building off of what we initially called our game detect technology we were able to develop technology that looked at anything that was happening on the network and now we can manage that individually.
Today our Killer Technology automatically detects games and optimizes games as the top priority. Latency sensitive chat or video chat, such as Skype is automatically handled as Priority 2. Web surfing, email, things like that are rated as Priority 3. File transfers, low-level system utilities things like that go at Priority 4. That’s the default theme that we shipped with but we could easily move to an enterprise client that we could optimize latency sensitive traffic out of the box for the enterprise. Yes, primarily when you see Killer Technology coming from Qualcomm Atheros it will be focused on games but that technology will be moving towards other mid-tier and enterprise technologies within the year.
DFC: Does this prioritizing require a software client or is that handled on-chip?
Sean: Hardware and firmware can do that without any user intervention. However, we have a software suite that runs on top of that in Windows that allows you to change the priorities and deeper network settings yourself. So, for example, with our visual bandwidth control application you can grab any application that is accessing Ethernet on your PC and throttle or increase the bandwidth allocated to that. So say, for example, that you are buying a game from a digital distribution service like Steam or Origin but you want to play a game while you’re waiting for that other game you download, what you can do is cut the total bandwidth allocation for Steam or Origin in half. Obviously it will twice as long to download the game but you’re guaranteeing that the download traffic won’t actually step on your gaming traffic, or your Skype traffic, or YouTube traffic, or whatever else that you’re looking to do.
DFC: Can Killer reduce latency to far distant servers?
Sean: As I have said many times, you can’t fix the Internet but we’re working on it. Killer concentrates on what is happening in your PC. People who run our control panel are surprised to see everything that they have on the PC that it is actually accessing the Internet. A lot of times Windows is looking for the same Internet access that your game is and Windows doesn’t ask nicely. They have a variety of reasons for doing so but you want to make sure that doesn’t step on your spell or your grenade or your sword swing or something like that. So by giving every application its own priority and its own identified stream and then putting games on top of that, you are guaranteed to knock out some of the little glitches in your networking traffic as long those glitches are confined to your PC.
DFC: Must developers optimize for Killer?
Sean: That was one of the very first questions we asked: “Is there an API or a set of best practices that we can encourage in game development to take even more advantage of our hardware?” The primary answer that we saw was developers don’t want to develop for another processor, especially when it comes to PC gaming. I don’t want to, by any means, say that there are people out there that don’t want to because obviously we are enjoying tremendous success on Snapdragon gaming. But what I will say is that a PC developer has enough to program already so we built Killer to work with everything. That philosophy was born out very well when we were able to identify and modify traffic from YouTube, or identify and throttle traffic from Netflix, automatically.
DFC: How are consumers finding out about Killer and how are they either purchasing it or getting access to it?
Sean: Our Ethernet solution is only available on Gigabyte motherboards curently. Our wireless solution is available from Alienware and about eight or nine other laptop providers here in the United States. How gamer are finding out about it is through our scrappy community efforts, social networking, as well as our pro gaming team sponsorship.
DFC: What’s next up in terms of game technology at Qualcomm?
Dave: We are actually working on quite a bit of stuff. I think that a lot of it is exciting. We just aren’t ready to announce or talk about it yet. By following some of the market trends and where stuff is going, some of what we have in the works might be obvious but we’re not quite ready to announce those publicly yet. There’s just so much going on.
