Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Specifically, an eight-core "Paxville" Xeon system clocked at 3 GHz versus the Core 2 Duo system clocked at 2 GHz, ceteris paribus.

This is assuming that the workload is effectively distributed across all cores in both systems.

Whicb of the two would pack more punch?

share|improve this question

closed as not a real question by Chopper3 May 22 '11 at 17:12

It's difficult to tell what is being asked here. This question is ambiguous, vague, incomplete, overly broad, or rhetorical and cannot be reasonably answered in its current form. For help clarifying this question so that it can be reopened, visit the help center.If this question can be reworded to fit the rules in the help center, please edit the question.

Paxville is a dual-core CPU, are you referring to a quad-socket system with four dual-core Paxville chips? – HampusLi May 22 '11 at 16:38
I suspect L2 cache size and locality of reference in the specific application is going to influence performance heavily. – Evan Anderson May 22 '11 at 16:51
Yep, quad-socket system is correct. Any chance that the two cores could beat out the eight? – Ryan Lester May 22 '11 at 17:02
up vote 0 down vote accepted

It's really tough to answer your question, because an even comparison between the machines isn't possible. One is going to have more sockets than the other.

Machines with more sockets are always optimized for different workloads than those with fewer sockets. They'll have more memory controllers and the I/O subsystems will be optimized for a lot more throughput. Consequently, they'll perform better at tasks that involve moving a lot of data through memory.

Machines with fewer sockets are much, much more power efficient. If you have a lot of them, you'll save money on electricity and you'll probably be able to pack more of them into a rack. They will also excel at any very tightly coupled computational task, where sharing caches is important, though such a workload is much more rare than the ones referenced above that are memory bound.

None of this matters at all if your workload is I/O bound (again assuming you have the same storage subsystem attached.) Even though your question stipulated that the workload evenly distributes across the cores, you still might find that I/O is your limiting factor, in which case you won't be able to keep those cores busy.

share|improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.