Aaron N. Tubbs bio photo

Aaron N. Tubbs

Dragon chaser.

Twitter Facebook Google+ LinkedIn Github

There’s an interesting
over at coding
horror where the argument is made that going from two to four cores doesn’t
provide much benefit, except for applications that are explicitly parallel, or
synthetic in nature (rendering, scientific, etc.).

For me (at work), the big advantage of multiple cores is that I can use Lotus
Notes and Visual Studio at the same time without causing a space-time rift.
Sometimes I can even have a web browser running at the same time as well.

So, no, I don’t expect that four cores is suddenly going to make Supreme
playable on any hardware currently
being produced. But, I stand a chance of being able to run a game, play some
streaming music in the background, keep a web browser running, and so forth,
without too much impact.

For me, gobs of cores isn’t so much about maximizing the performance of a
single application, but maximizing the usability of a computer under a typical
work load. Whether it’s games or work, I very seldom have one application that
I run, where I want to dedicate 100% of my computer’s CPU, memory, and IO.
Instead, I’m running tons of stuff, and I’d like to have it all perform well.

While I have no doubt that there will be games that can effectively utilize n
cores at some point in the future, I think there are also some big issues
before that becomes realistic; I think John Nack’s article on multi-core
performance in
a good argument in this direction — cores aren’t everything.