Opening The Worm Can

[Jabo]

Hey chaps,

I've just come into the (legitimate) possession of an MSI P7N Platinum mobo based on the nVidia 750i chipset, it's also got an intel quad-core processor (Q6600) installed. The question is this...Do I swap this board/processor with my EP45-DS3 Gigabyte & its Core 2 Duo or do I swap CPUs or do I do neither?

[Kimosabi]

Depends on how fast your Duo is and what you do with your computer. To somewhat ready that rig for SoW, I'd install the Q6600 on the P45 board.

[Jabo]

Thanks Kimo

Both chips are rated at 2.4GHz, so there's not much difference other than the additional cores. I think I'll benchmark the rig now, then chuck in the 6600 and benchmark it again.

I don't think the MSI board gives me anything I don't currently have on the Gigabyte other than eSATA (which I don't use) and half the number of USBs on the back (which I do) so assuming the 6600 is good then that board and my C2D will be the basis of a new rig for the GF.

Jabo

[BadAim]

Yeah, I agree, I think the p45 is probably a better chipset. One thing you might want to do is make sure your bios is updated to support quadcore, it probably is good, but worth a few minutes on the Gigabyte site to check it out. I'm quite sure that you'll see little performance gain in IL2, but then again having extra cores for handling the background tasks might help a bit.

[Kimosabi]

The P45 is a better chipset all around(never get x16/x16 from it though but that doesn't really matter unless you're on SLi GTX480's/CF 5970's) but if you're running more than one gfx card and it's Nvidia, you'll need the 750i chipset to run SLi. Unless you know how to SLi hack a P45. The downside of a 750i chipset, from an overclockers standpoint, is that it doesn't fancy a Quad all that much.

[Jabo]

Well, the 6600 gives me a whole 5.6% improvement in Passmark (whoopy-doos), but here's an interesting thing, the CPU test reports a 30% improvement which is probably fair, but all the other component tests (memory, gfx etc) give a worse score than with the old CPU in - up to a whopping 20% (Worse!!!) for the 2d gfx test. Why is this? (Oh and Kev, we were wrong, it's not power related)

Jabo

[JensenPark]

mmmmmmmmm....worms...

[Jabo]

lol

[Kimosabi]

Well, the 6600 gives me a whole 5.6% improvement in Passmark (whoopy-doos), but here's an interesting thing, the CPU test reports a 30% improvement which is probably fair, but all the other component tests (memory, gfx etc) give a worse score than with the old CPU in - up to a whopping 20% (Worse!!!) for the 2d gfx test. Why is this? (Oh and Kev, we were wrong, it's not power related)

Jabo

Probably because you have a well balanced system which doesn't really demand more processing power. Duo's are faster chips clock-for-clock compared to Quads, and since most games today doesn't really utilize more than one-two cores, a Duo is still recommended for playing games like IL-2. Just because it's a spiffier proc. The best way to measure your rigs gaming potential though is to run it through 3Dmark 06 or 3Dmark Vantage.

Oh, SoW will utilize multiple cores.

[Jabo]

Point taken m8, but why have other component tests now got worse scores

Jabo

P.S. The new proc is not really for IL2, but I have an eye on other games and (of course) SoW - Two weeks apparently.

[Kimosabi]

What's your total score?

[Jabo]

1215

[Kimosabi]

If that score is better than the total score with a Duo I wouldn't think too hard about it. CPU score counts like 29% or something in Passmark anyway and I wouldn't trust the program that much. It read my overclock to 3.9gHz when it's only 3.75. If the rig feels good, you get good framerates and it runs smooth, you're all good m8.

It is strange though. My best bet is still that the Quad is not as fast as a Duo clock-for-clock. I could make a posting about it at OCN to find out more but then I would need the full spec list, from head to toe. My rig overclocked scores 2278 in comparison. Q9550 and Gigabyte P45 UD3P with a 5870.

[Jabo]

Yep, the Duo scores 1149 against the Q6600 (1215). The bizarre thing is that the processor test shows a 30% improvement in performance, but everything else takes a hit. If nothing else changed I should be seeing a 10% overall improvement, rather than the 5 and a bit I have. It's not so much the processor which is intriguing me, but why everything else is worse.

Jabo