PS3 Developers and Multi-core Processors
One of the big things about the PS3 (and Xbox 360 for that matter) is that it has a multicore processor. One main PowerPC core, and 7 Synergistic Processor Elements (or SPEs). An SPE is basically just another core that’s geared more towards a certain kind of computation. So the Cell chip has 8 cores.
And, well, you know, this means the end of all time as we know it.
You’ve read the doomsayers as well as I have. Game programmers won’t be able to take advantage of all those cores. Those extra cores should be tossed in the trash just like apple cores. The first generation of next generation games will all just use the main core and ignore the others. Programming has gotten much too complicated. Blah blah blah. If you’ve been reading this website and others, you’ve seen it.
And while we read all this and think about how terrible this next-gen mess has turned out to be, Apple introduces a computer with to processors each of which has two cores. That’s four cores total. Of course, for an application to really scream in this scenario it has to be multithreaded up the apple core. But do you see huge protests? Big hot shots in the Apple development community complaining about how terrible this is and that the sky is falling? Any of that? No. It’s just the next step. A little more work. Great benefits, if you choose to do the work. Not if not.
So PS3 (and Xbox 360) game developers will have to learn how to write multithreaded applications. Deadlocks, mutexes, condition variables, semaphores. You know the drill. Programmers have been doing this for literally decades. It’s tricky, granted. But I think those game developers are smart folks. They’ll do it, and we’ll reap the rewards.