How Many Processor Cores Are Enough?
|
From: Jon Forrest on 27 Sep 2006 17:10 Today I read that we're going to get quad-core processors in 2007, and 80-core processors in 5 years. This has got me to wondering where the point of diminishing returns is for processor cores. We've all seen those bench marks in which the same test is run on a system with 128MB of RAM, then 256MB, then 512MB, ... What usually happens is the the first increase results in a big improvement, the next increase in a smaller improvement, and so on. At some point, the size of the improvement doesn't justify its cost. I suspect we'd see the same kind of results if we could increase the number of processor cores. The big difference here is that additional processor cores only help if there's work for them to do. Modern operating systems use multiple processes and threads but I wonder how many processes/threads are runnable at any one time in a general purpose computer running general purpose jobs. Where do you think the point of diminishing returns might be? Jon Forrest
From: Casper H.S. Dik on 27 Sep 2006 18:04 Jon Forrest <forrest(a)ce.berkeley.edu> writes: >Today I read that we're going to get quad-core processors >in 2007, and 80-core processors in 5 years. This has >got me to wondering where the point of diminishing returns >is for processor cores. Sun has been shipping 8 core CPUs since, I think, late last year. It all depends on the bandwidth. (Which means it ain't a pretty picture for Intel as long as they keep the FSB) Casper -- Expressed in this posting are my opinions. They are in no way related to opinions held by my employer, Sun Microsystems. Statements on Sun products included here are not gospel and may be fiction rather than truth.
From: Joe Seigh on 27 Sep 2006 18:23 Jon Forrest wrote: > Today I read that we're going to get quad-core processors > in 2007, and 80-core processors in 5 years. This has > got me to wondering where the point of diminishing returns > is for processor cores. > > We've all seen those bench marks in which the same test is > run on a system with 128MB of RAM, then 256MB, then 512MB, ... > What usually happens is the the first increase results in > a big improvement, the next increase in a smaller improvement, > and so on. At some point, the size of the improvement doesn't > justify its cost. That's called scalability. > > I suspect we'd see the same kind of results if we could > increase the number of processor cores. The big difference > here is that additional processor cores only help if there's > work for them to do. Modern operating systems use multiple > processes and threads but I wonder how many processes/threads > are runnable at any one time in a general purpose computer > running general purpose jobs. > > Where do you think the point of diminishing returns might > be? > It all depends. The industry has a real chicken and egg problem. Applications won't start adapting to the new architecture for a while yet. And what the new architecture will actually end up being isn't real clear. It could be shared memory, NUMA or otherwise. It could be based on message passing. Or something else entirely. -- Joe Seigh When you get lemons, you make lemonade. When you get hardware, you make software.
From: Elcaro Nosille on 27 Sep 2006 18:26 > Sun has been shipping 8 core CPUs since, I think, late last year. Yes, but on server-systems you've got an appropriate workload.
From: Dennis M. O'Connor on 27 Sep 2006 19:01
"Jon Forrest" <forrest(a)ce.berkeley.edu> wrote ... > Today I read that we're going to get quad-core processors > in 2007, and 80-core processors in 5 years. This has > got me to wondering where the point of diminishing returns > is for processor cores. It depends on your application. Examples: A web server like Apache can effectively use a lot more cores than an old DOS game. But future games on the PC may be able to effectively exploit every core you can provide them. -- Dennis M. O'Connor dmoc(a)primenet.com |