As Ben Kuchera is gearing up to report at Opposable Thumbs, one of the themes for this year's GDC is mobile gaming—and iPhone 3D gaming, in particular, is red-hot. And for good reason: some of the demos so far are amazing (I'll let Ben tell you about them as the embargoes lift), and it's nuts to think that this level of 3D performance is taking place on a tiny mobile device. But as I watch the mobile 3D action this week, one fact above all others bowls me over: this is a mass-market 3D revolution, with an installed base of millions—and it involves neither Intel nor Windows.
I've covered at some length the ARM offerings in this space, and why Intel won't have a shot at a real mobile phone form factor until sometime after transitioning to 32nm, but I've paid less attention to the software side of this equation. At a GDC session yesterday by the Khronos Group, a broad industry consortium working on the OpenGL, OpenCL, and other GPU-related APIs, I was surprised at just how little sway Microsoft has in the mobile 3D arena.
Sure, Microsoft has D3D Mobile, but the Windows Mobile platform also supports OpenGL ES—and if you're developing a 3D game right now for the mobile market, why wouldn't you just use OGL? Besides, Windows Mobile isn't exactly a hotbed of gaming activity. No, the mobile 3D revolution is, for now, an iPhone phenomenon, though it's entirely likely that this won't last.
At some point, Microsoft will have to get its platform act together, and hopefully in a way that ties in meaningfully with the Xbox Live ecosystem. If anyone could pick up the mantle of the original Nokia N-Gage concept and run with it, it's Redmond.
Palm, too, is a potential mobile gaming contender. The upcoming Pre is powered by a substantial media SoC, and I'm anxious to see what game developers will do with it. Pre also has a potential advantage on the control front, since it sports a special gesture region at the bottom of the screen.
The main issue with both WinMo and the Pre, though, is that they're focused squarely on business productivity and communication use cases. The Pre in particular, as the first mobile platform that presumes the existence of the cloud and the social web, puts most of its eggs in the networking basket. So much so that its software stack may have too much overhead to support processor-intensive 3D games. Palm may have to give game developers a separate set of APIs that put them closer to the hardware, in order to save on battery life and enable sufficient performance.
Even though Windows and Intel will be late to the mobile 3D party—and they'll probably each show up stag instead of as a couple—this doesn't necessarily put them at a crippling disadvantage. Mobile apps, games, and devices are fairly ephemeral compared to their desktop counterparts, demanding much less in the way of investment and commitment from either developers or end-users. What this means is that legacy issues are minimal, since as a developer you can just make the next round of games largely from scratch for the latest and greatest phone. So if and when Intel and Microsoft are ready to jump into mobile 3D gaming, if they can build it, developers and users will probably still come.
source
