PCMark2005 Tweaks

[Kunaak]

What are the Tweaks to PCMark2005?

I read some posts on Futuremark, but honestly, the posts there are so muddled with gossip and half information that it's hard to weed out the good info, from the bad... so I just kinda gave up there.

from what I saw, theres a tweak about editing your Boot.ini by adding /usepmtimer to the lines... I did it right, but I got not luck from it.

I also saw some info about the AMD64 driver offering some big point increases, but I got zero luck with that too.

theres scores that are done with PC's that are signifigantly slower then my own PC, that are over 1000 points ahead of my own PC and I can't figure out what they are doing to get that high.
cause theres no way, a 2.7 ghz A64 3800 X2 with no raid, and a X800XL should be wasting my own PC of a 3.07 ghz Opteron 170 dual core, with 4 hard drives in raid, and 7800 GTX.
I am faster then that PC in all aspects, yet the scores are WAY off, compared to what I see others getting.
now I know its tweaking, I just can't figure out what they are doing to get that high.

[[XC] moddolicous]

Hard drive speed is really important in Pcmark 05. I think the guy in first has seagate 15k cheetahs in raid. Mem timings also play an important part.
Here, this might help:
http://www.tweakhound.com/xp/xptweaks/tweakbench.htm

[IluvIntel]

What are the Tweaks to PCMark2005?

I read some posts on Futuremark, but honestly, the posts there are so muddled with gossip and half information that it's hard to weed out the good info, from the bad... so I just kinda gave up there.

from what I saw, theres a tweak about editing your Boot.ini by adding /usepmtimer to the lines... I did it right, but I got not luck from it.

I also saw some info about the AMD64 driver offering some big point increases, but I got zero luck with that too.

theres scores that are done with PC's that are signifigantly slower then my own PC, that are over 1000 points ahead of my own PC and I can't figure out what they are doing to get that high.
cause theres no way, a 2.7 ghz A64 3800 X2 with no raid, and a X800XL should be wasting my own PC of a 3.07 ghz Opteron 170 dual core, with 4 hard drives in raid, and 7800 GTX.
I am faster then that PC in all aspects, yet the scores are WAY off, compared to what I see others getting.
now I know its tweaking, I just can't figure out what they are doing to get that high.

Did that user who had the A64 3800 X2 have Asus mobo with PCI express ?

[.sentinel]

Why would an Asus board matter? Also do you guys have Windows 2000 tweaks?

[Circaflex]

because i think there was some blitch with asus mobos in pcmark05 but i thought it was intel that glitched but i dunno.

[IluvIntel]

Why would an Asus board matter? Also do you guys have Windows 2000 tweaks?

Have a look at this thread (http://www.xtremesystems.org/forums/showthread.php?p=1085621#post1085621) ;)

[Vapor]

It's only 925X(E) boards (maybe some 955X, not sure)....predominantly ASUS. A64s do not have this problem. Sorry for being confusing back then.

[Vapor]

Kunaak....those two tricks you listed do nothing (did nothing for me at least--and my score's not in deficit), people reported placebo.

If you would, could you please post a compare link so I could see exactly where you're falling behind?

[Kunaak]

I've been tweaking my OS for a week.

I also said in my first post, that the tricks listed, did nothing for me.
in fact, I have yet to find a single tweak, or driver, or anything else that actually helped my score at all.
only obvious things seem to do anything so far.

higher clocks.
higher HTT.
higher bandwidth...
play with drivers.
play with your raid stripe.

anyways, I have learned on major thing however.

CPU means almost nothing in this benchmark, videocards seem to have more effect then the CPU speed, but what the real power comes from, is hard drive speeds.
all the tops people are running SCSI or crazy raid set ups.
hard drives mean everything in this benchmark.

right now, my best score is 7249, at 3.07 ghz with a Opteron 170.

http://service.futuremark.com/compare?pcm05=192267

I am not happy with that score at all.
theres alot of people who are running signifigantly slower PC's then mine, and are scoring much higher then mine, due to thier raid and SCSI set ups, cause thats where they always beat me, especially in windows start up, and hard drive general use.
I destroy them in all other areas, sometimes by as much as 4x in some areas, like videocards, yet it does almost nothing... I simply can't make up for the difference with raw CPU speed or videocard OC's... cause hard drives mean everythin in this benchmark, and thats the biggest weakness in my PC.

2 x Sata2 drives in raid zero, is nothing, for the top 20.
I would say I probably have the slowest hard drive scores of anyone in the top 20 of PCMark2005 ....

[Vapor]

Yeah, man, my single Raptor beats up on your HDD setup and I score nearly the same score with only one GFX card and 300 less MHz on my CPU.

Those HDD scores are very low....you beat me in just about everything else (except transparent Windows because I haven't run the new, less tweakable version yet and 64 lines [my GFX memory is clocked 100MHz higher]).

http://service.futuremark.com/compare?pcm05=73291

[Kunaak]

strange, 7188 would put you at spot number 16 on the top 20 for PCmark2005... but your not on the list.

anyways, your running a single raptor?

[Vapor]

I ran the old version and haven't even dl'd the updated one yet (haven't really cared). My score would probably drop ~200 from the lack of the transparent windows test tweak but probably gain back 100 of that with the improved clocks and new drivers for my GTX.

Yeah, single Raptor. Also have RAMDrive scores, but it was a software drive so they don't count. Got nearly 11k with the same setup though (shows how much a HDD boosts it--my RAMDrive wasn't even as fast as they get either [not even close, TBH]).

Anyway, there's a good deal or two, IIRC, in the FS forum for a Raptor, might want to check there.

[d]g[ts]

Kunaak It seems that nvraid is the issue, but can be helped by disabling the read disk cahe on the drives. if ya do that you should see an improvement on the HDD tests.

this issue of poor performance was killing me as i was scoreing better in hdd tests without raiding the drives.