: : : Core 2 Duo Scaling in Gaming

Core 2 Duo Scaling in Gaming - PAGE 1

- Thursday, January 24th, 2008 Like Share






Get updates when we publish new articles

Comments

Sort by date: ascending descending
0 thumbs!
^
Iceguy2003 Jan 25, 08
Interesting, although I've always thought (from my experience) that a C2D @ 2.2 - 2.4ghz was plenty of power. I've had the same kind of minimal gains, and have just recently brought my CPU down to 2.5ghz. I don't see any point in going over 2.5ghz. 2.5ghz and keeping the memory clocked high as possible seems best to me.

Although, I would have liked to have seen how Athlon X2 CPUs would have held up at the same speeds, and see when they quit showing performance increases, too. But don't take that as an insult, I'm just stating that it has me curious.
0 thumbs!
^
calvary1980 Jan 25, 08
I thought a Quad Core should of been in it you had nothing to compare.

it was well written but not informative. Thanks Kevin.

- Christine
0 thumbs!
^
Darkness Flame Jan 26, 08
Hmm ... I both can, and cannot agree with this article. I mean, certainly, with the exception of anything pre-2007, most games seem to not benefit as much from higher clock speeds, especially above the 2.2-2.4 GHz range.

However, what I can't agree with, is that you only really compared in terms of clock speed. For example, comparing an E6420 and an E6550, there can be about a 5-10 frame difference.

Their are only two main differences in those cores, and as you've show, 200 MHz clock difference won't give you that. However, the FSB difference, I think can. Basically, try comparing two 2.4 GHz processors. Clock one at 200x12, and one at 400x8. If they're both using DDR2 800 ram, (which seems common for gamers now a days, especially since the price drop), then I think that you would find that you could indeed get higher results.

Basically, I personally have the theory that it's not so much the processor has an excess clock speed in general, but rather, it has an excess clock speed compared to the FSB. I mean, when I've seen an E6600 compared to an E6850, there could be up to 10 frames difference. Mind you, any of these tests that I have talked about were on older games (IE: UT 2004, Prey, Quake VI). However, if you were to do a test like that with current games, while it might not be the same, I do still believe that you could find a difference.

As a note, while typing this comment, I was also browsing, and came across a CPU and GPU performance comparison for UT 2007. One test was between 1066 MHz FSB, and 1333 MHZ (as well as between dual and quad core processors.) As shown, the higher 1333 MHz usually got at least 5 frames higher. Can be found here
0 thumbs!
^
kspiess Feb 4, 08
Thanks for much for the comments guys, appreciated and noted.

There was many ways of going about this article and its great to get some feedback on it.

For a future article, I think I would like to take a look at the affect of L2 cache size on gaming performance perhaps.

Iceguy -=> Same for me, pretty much. I cut my overclock of my E2180 by 300 MHz just to extend the life of the CPU a bit. Having some X2 benches for comparison would have been good.

Calvary-=> Thanks. I was going to do a chart up with a Penryn, but changed my mind, just wanted to keep it about Core 2s.

Darkness Flame-=> Good points. You are right about the FSB thing -- FSB is more of a determinant on performance than your clock speed. I should have talked about my FSBs and multipliers -- kind of surprised I didn't mention that, now that I look back on the article. FYI I kept a 8x multiplier and raised the FSB by 25 at for each 200 MHz 'step' in the charts. Maybe I'll go back and add that...
0 thumbs!
^
Guest Reader Mar 29, 08
Jeez, did anybody proofread this article before publication? It contains a large number of grammatical errors, such as there/their, your/you're, affect/effect, missing words, redundant words, etc.
Sort by date: ascending descending
Add your comment:
Name *:  Members, please LOGIN
Email:  We use this to display your Gravatar.

Sign in with
Comment *:
(0.0719/d/web7)