Four Thoughts On Quad-Core . . .


1) Even though the first desktop chips are supposed to show up in early next year, this isn’t even going to get a fighting chance to go mainstream until the CPU manufacturer gets to 45nm, simply due to size. Since it looks like Intel will get there first, they’ll have the headstart here, just like they have the manufacturing edge in dual-core with their 65nm process.

2) It’s becoming apparent that outside of relatively few, usually professional programs that benefit greatly from parellization, any parallelization that’s going to be done will either get done by a compiler/precompiler, or it won’t be done at all.

3) Just what is Vista/Linux/MacOS X going to do with two CPUs, much less four? Will the OSs actually be able to do some smart things with them, or not?

4) Am I the only person who is being to wonder what actual, average-person purpose quad-core processors will serve other than to give AMD and Intel a chance to keep their prices up? Yes. I suppose that eventually, some use can be found for the average schmo for four processors (“Processor one runs your HDTV, processor two runs your DVD, processor three runs your thermostat, and processor four does everything else,” but aren’t some people, whether they’re in the Third World, or in the inevitable “put a PC in a phone” business going to say, “If you can give me a quad-core for $200, why can’t you give me a dual-core for a hundred, or single-core for $50?”

Ed


Be the first to comment

Leave a Reply