VMware Cloud Community
iaco
Contributor
Contributor

comparing Mhz from 2005 to 2011

i know i may asking for the holy grail here, but i'd like to have some opinions on that.

i'm going to plan an hw refresh for a farm built way back in 2005, based on Xeon MP 3.0 Ghz processors, with newer blades with some X5660 processors.

looking at the vcenter statistics i can see the actual farm is using something like 50Ghz as an average, and has a physical to virtual cpu ratio of about 1.5.

by a rule of thumb, i may say that on newer processors, doubling the p2v cpu ratio won't cause any problem, even if i don't have any evidence to prove it.

speaking of Mhz instead, i'm trying to find out how a 2011 core compares to a 2005 one.

even if X5660 cores are running at 2.8Ghz, i don't think they're "slower" than Xeon MP ones.

is a standard benchmark like specint2006 a viable tool to compare cores "speed"? so may i say that 25Ghz from a 2011 X5660 are equivalent to 50Ghz from a 2005 Xeon?

Reply
0 Kudos
6 Replies
AndreTheGiant
Immortal
Immortal

Frequency is only one aspect. And you're right... isn't grow due to some physical/thermal limits.

But you have consider also the cache (L1, L2 and L3), core architectures, bus speeds, ...

Andre

Andrew | http://about.me/amauro | http://vinfrastructure.it/ | @Andrea_Mauro
Reply
0 Kudos
lanamarkinc
Enthusiast
Enthusiast

Lanamark Suite Services Edition normalizes workloads between source and target CPU architectures instead of using MHz. If you submit an online inquiry referencing this thread, we will respond to you promptly.

Reply
0 Kudos
idle-jam
Immortal
Immortal

Platespin Recon also normalized the Mhz to the latest technology ..

Reply
0 Kudos
iaco
Contributor
Contributor

i've fiddled a litte with the published spec.org CINT2006 benchmark results

here is a sample

xeon 3.8Ghz         11.4
e7330 2.4Ghz        17
e7458 2.4Ghz        20
x5560 2.8Ghz        35

Reply
0 Kudos
FredPeterson
Expert
Expert

Based on your 50GHz and 1.5 vCPU/pCPU ratio, I'm going to take a WAG and say that'll drop to something like 18GHz.  You will see an absolutely incredible amount of performance gain.

Normalizing those SPECint figures doesn't mean much because it assumes a linear scale up factor.  Newer processors are much more about scale-out: cache speed, core count, memory speed, die size etc.  All of these factor into why a slower CPU can still be way faster.

So while it appears the higher end CPU is only about 3 times faster - thats only on a single thread that it is 3 times faster.  Consider your Xeon MP 3.0GHz was only capable of processing 4 threads at once, a x5660 will do 12+

Reply
0 Kudos
iaco
Contributor
Contributor

i've stumbled upon this blog post that seems to came to the same conclusion

http://it20.info/2011/06/the-cloud-and-the-sunset-of-the-ghz-based-cpu-metric/

Reply
0 Kudos