chrisng
01-26-04, 07:50 PM
Okay, here's the situation...
I have 2 XP cpus, a 2200+ thoroughbred B and a 2500+ Barton..
Unfortunately both of these chips are locked :(
Anways, I decided to do some benchmarking today, and the results kind of surprised me..
I have the 2200+ overclocked to 2.1ghz (160fsb) and the 2500+ @ 2.09ghz (190fsb) ..
the 2200+ was detected as a 2600+ and the 2500+ as a 2800+ .. I fully expected to see the overclocked 2500+ to outrun the 2200+ because of the higher PR rating and larger cache...
I ran 4 Benchmarks, 3dmark03, PCmark04, Sandra and Aquamark..
The results we as follows:
3dMark03:
2200+ @ 2.1 - 3761 3d marks 488 cpu
2500+ @ 2.09 - 3779 3d marks 530 cpu
Although there was a diffrence in cpu marks, the overall 3dmarks were very close
PCmark04:
2200+ @ 2.1 - 3735 PCmarks
2500+ @ 2.09 - 3723 PCmarks
Although very close marks, the 2500+ benched slower
Sandra:
Arithmetic benchmark:
2200+ @ 2.1 - 8243 mips 3407 mflops
2500+ @ 2.09 - 7965 mips 3289 mflops
Multimedia benchmark:
2200+ @ 2.1 - 19855 it/s 20388 float it/s
2500+ @ 2.09 - 19169 it/s 19675 float it/s
Sandra benches the 2200 quite a bit higher.. hmm....
Aquamark:
2200+ @ 2.1 - 30092 (cpu : 6598 gfx: 3897)
2500+ @ 2.09 - 30494 (cpu : 7225 gfx: 3864)
So judging the benchmarks of the 2 overclocked cpus, there doesn't seem to be much of a diffrence.. But I thought the extra 256k cache of the Barton core should given the 2500+ an edge.. I've also heard that a higher FSB shows more of an improvement over a higher multiplier.. well in this case, 13x160 benches faster than 11x190!
All the tests we conducted on the same motherboard with no other changes to any settings or configuration other than popping in the CPUs and changing the fsb.. Another thing I noticed was that the 2200+ overclocked runs at 41C idle, 49C full load, while the 2500+ sits at 45C idle, 52C full load.. wern't the Bartons supposed to run cooler?
perhaps I'm looking at this all wrong and relying too heavily on these benchmarks to make my conclusions as to which cpu is better.. maybe someone out there could shed some light as to why my 2500+ barton would be a better chip to keep around then the 2200+ ...
I have 2 XP cpus, a 2200+ thoroughbred B and a 2500+ Barton..
Unfortunately both of these chips are locked :(
Anways, I decided to do some benchmarking today, and the results kind of surprised me..
I have the 2200+ overclocked to 2.1ghz (160fsb) and the 2500+ @ 2.09ghz (190fsb) ..
the 2200+ was detected as a 2600+ and the 2500+ as a 2800+ .. I fully expected to see the overclocked 2500+ to outrun the 2200+ because of the higher PR rating and larger cache...
I ran 4 Benchmarks, 3dmark03, PCmark04, Sandra and Aquamark..
The results we as follows:
3dMark03:
2200+ @ 2.1 - 3761 3d marks 488 cpu
2500+ @ 2.09 - 3779 3d marks 530 cpu
Although there was a diffrence in cpu marks, the overall 3dmarks were very close
PCmark04:
2200+ @ 2.1 - 3735 PCmarks
2500+ @ 2.09 - 3723 PCmarks
Although very close marks, the 2500+ benched slower
Sandra:
Arithmetic benchmark:
2200+ @ 2.1 - 8243 mips 3407 mflops
2500+ @ 2.09 - 7965 mips 3289 mflops
Multimedia benchmark:
2200+ @ 2.1 - 19855 it/s 20388 float it/s
2500+ @ 2.09 - 19169 it/s 19675 float it/s
Sandra benches the 2200 quite a bit higher.. hmm....
Aquamark:
2200+ @ 2.1 - 30092 (cpu : 6598 gfx: 3897)
2500+ @ 2.09 - 30494 (cpu : 7225 gfx: 3864)
So judging the benchmarks of the 2 overclocked cpus, there doesn't seem to be much of a diffrence.. But I thought the extra 256k cache of the Barton core should given the 2500+ an edge.. I've also heard that a higher FSB shows more of an improvement over a higher multiplier.. well in this case, 13x160 benches faster than 11x190!
All the tests we conducted on the same motherboard with no other changes to any settings or configuration other than popping in the CPUs and changing the fsb.. Another thing I noticed was that the 2200+ overclocked runs at 41C idle, 49C full load, while the 2500+ sits at 45C idle, 52C full load.. wern't the Bartons supposed to run cooler?
perhaps I'm looking at this all wrong and relying too heavily on these benchmarks to make my conclusions as to which cpu is better.. maybe someone out there could shed some light as to why my 2500+ barton would be a better chip to keep around then the 2200+ ...