Is it better to have a higher FSB or Multiplyer?

[KhoiFather]

I'm wondering because I thought having a higher fsb would be better. Check this, I did 178 x 10 = 1780MHz and scored like a 3044 on 3DMark03 but when I put it on 166 x 10.5 = 1743MHz, I scored 3055. I thought I would get a higher score since I have more bandwidth and what not. I also ran PCMark2002 and same thing happened.

 

[BlueWeasel]

Well, in theory, it's better to have a higher FSB, but the performance increase in some applications may not be as evident as in others. Also, there's not much difference between 1780 (178 FSB) and 1743 (166), and that's why you are getting similar scores. Given 166 FSB and 200 FSB, there'd likely be a more significant difference in scores.

Are you sure that your chip and RAM are running insync?

 

[KhoiFather]

Yeah both the cpu and ram is insync

 

[myocardia]

Khoi, see what the difference is when running it at 10x200, or even better, 10x210. BTW, your difference between the first two tests was less than 1%. Run it 3 more times, and it's likely to change every time, at those two fsb's. I can tell you from experience, though, that 3dMark likes cpu speed more than a high fsb. Try benchmarking with something other than a 3dMock app, like SuperPi, and I think you'll notice more of a difference.

 

[MegaWorks]

higher FSB always

------------------------------------------------------
AMD Athlon XP T-Bred B 1700+ @ 2.3GHZ (1.775v) 400FSB = 3200+
Abit NF7-S V.2 (nforce 2 400U)
Corsair TwinX XMS 3200LL 512MB @ 2-3-2-6 (2.6v)
Sapphire Radeon 9500 Pro 128MB
Antec TrueBlue 480W
Thermalright SLK-947U with 92mm Vantec Tornado @2800RPM
Maxtor 91531U3 15GB
2x Western Digital WD400BB 40GB serials
Antec PlusView 1000AMG
Cambridge SoundWorks MegaWorks THX 550 5.1