Messages in this thread | | | Date | Mon, 24 May 1999 19:42:08 -0400 (EDT) | From | Stephen Frost <> | Subject | Re: andrea buffer code (2.2.9-C.gz) |
| |
On 23 May 1999, Stefan Monnier wrote:
> >>>>> "Andrea" == Andrea Arcangeli <andrea@suse.de> writes: > > Impacting runtime performances to SMP scale better, is going to lose > > according to me. I think that if you want to SMP scale better it worth to > > pay _only_ with memory wastage and _not_ with wasted CPU cycles. > > The rule generally goes in the other directions for fairly obvious reasons: > the amount of memory-resources per CPU is usually lower on an SMP so you > want to save memory more than you want to save CPU.
Hmm, I probably have NO clue what I'm talking about, but it would seem to me that in general, SMP machines have more memory than uni-processor machines because in general they play a different role. A uni-processor machine is going to have 32M or 64M probably, whereas a dual-proc machine is going to have 128M+ most likely. Why? Because if you're going dual-proc there is usually a reason, perhaps it's for a server, or a high-end workstation or something, but there is usually a reason, and that reason usually follows that the machine will have more memory. Also, memory is (relatively) inexpensive compared to processors. Which also makes me think that dual processor machines are going to have more memory than uni-processor machines. How much memory are we talking here, anyway? And how many processor cycles? :)
Stephen
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |