PCMark05 Results ...

[eva2000]

I'm not a member and thus can't see that page....do you happen have a compare link of his? I'm in SERIOUS doubt that RAIDing a set of Raptors together could boost it 1000 points, especially from all the 100-400 point boosts I've seen.

that is my post on OCAU forums (nice forum to join )

http://forums.overclockers.com.au/sh...3&postcount=48

yup single 120GB 1200JB = 5102 max for my FX-57, with 4x 80GB Hiatchi 7K80 SATAII NCQ raid 0 with FX-57 = 6,112

hdd is biggest difference you will find with pcmark05, even single 74GB Raptors give 300-400 more pts than regular SATA single drives due to their speeds

i have a pcmark05 calculator spreadsheet at http://i4memory.com/showthread.php?t=787, go through to virus hdd, general hdd usage and xp startup time changes and see diff in scores

FX-57
PCmark05 = 6,112
http://service.futuremark.com/compare?pcm05=52639

[Vapor]

Hmmmm, I was under the impression that RAIDing (even if they're 4 faster drives) simply wouldn't add that much (based off my RAMdrive results). Do you have a compare link to before you had the RAID0 array? I'd like to see if it gains anything in tests other than the labelled HDD tests.

[eva2000]

nah don't have that exact compare url for pcmark05 but i have one with single 160GB Seagate 7200.7 NCQ drive of 5,118 but that was with single 7800GTX http://service.futuremark.com/compare?pcm05=46734

[Vapor]

Good enough, I'll ignore the PS and 64 lines results as they're strictly graphics card related anyway.

[Vapor]

Hmmmm, my analysis shows that nothing gains other than the three HDD tests. And to a large degree, I was right about the Raptors in RAID, they won't gain the same 1000 points. Why? Look at eva's Virus test....seriously slow in the 5118 run, my Raptor gets 3x that score but RAIDed Raptors don't go much past the 60MB/s he got with his RAID setup.

[eva2000]

that 60MB/s is 4x 80GB Hitachi 7K80 SATAII NCQ raid 0 no raptors

[Vapor]

I know...I get 48MB/s with 1 and RAIDed Raptors only get about 60MB/s....seems to level off (my RAMdrive only got 11MB/s or some crazy like that!). That's where the gain is to be found in RAIDing if you don't already have a good score. That seems to be why Raptors simply don't gain much by going to RAID.

[shoe]

http://service.futuremark.com/compare?pcm05=73309

8893, 2nd in the world.

[mike]

http://service.futuremark.com/compare?pcm05=73309

8893, 2nd in the world.

NICE.....

[SamHughe]

Here's my updated score: 5136 (broke 5k with mild overclocking)

SamHughe

Wait a sec! My new rig scores only 300 more (5462) than my old rig? Great!

[Charles Wirth]

Shoe's score has a 2GB ramdisk and still scored poorly.
I setup a ramdisk to play around with and scored 16K +

I am running 4x raptors now and playing with stripe size and registry settings for cache write back and size.

I can get a good startup and general usage at the cost of other scores going down. Something is screwy with PCM2K5 as the raptors should score much better in virus scan. I can get teh virus scan up but other scores go down as a result.

[t800]

playing with stripe size and registry settings for cache write back and size.

What values exactly do you change in registry, to tweak cache write back?

[davidw8818]

i got 4790 points with a 2.9ghz sandy ill post pics later

[boshuter]

Heres my best score so far.... 7344

Good enough for 8th on the orb, overall This is with water cooled cpu, gpu on stock air/volts, and single raptor hdd.

[eva2000]

Shoe's score has a 2GB ramdisk and still scored poorly.
I setup a ramdisk to play around with and scored 16K +

I am running 4x raptors now and playing with stripe size and registry settings for cache write back and size.

I can get a good startup and general usage at the cost of other scores going down. Something is screwy with PCM2K5 as the raptors should score much better in virus scan. I can get teh virus scan up but other scores go down as a result.

it's cause raptors only support TCQ an dnot NCQ i think

here's my latest with my FX-57
pcmark05 = 6,149
http://service.futuremark.com/compare?pcm05=87280

[boshuter]

it's cause raptors only support TCQ an dnot NCQ i think

here's my latest with my FX-57
pcmark05 = 6,149
http://service.futuremark.com/compare?pcm05=87280

I stuck a 73gig 15k4 Seagate scsi in running on an LSI U160 controller and ran into the same problem Fugger is talking about. I got a lot better Windows loading score and general usage score, but the virus scan was way down. Overall the single 74gig raptor outscored the 15k scsi by about 60pts. The scsi would have killed it if I could figure out how to get the virus scan up with it.

[g8ts]

IS this the same test you are all running
Thanks for the help,looking for a place to get real feed back on scores.

http://service.futuremark.com/compare?3dm05=1280176

[D_o_S]

it's cause raptors only support TCQ an dnot NCQ i think

here's my latest with my FX-57
pcmark05 = 6,149
http://service.futuremark.com/compare?pcm05=87280

Raptors support TCQ, whereas SATA II drives support NCQ (different technology, you need a special controller for TCQ unless I am mistaken).

[BigStan]

IS this the same test you are all running
Thanks for the help,looking for a place to get real feed back on scores.

http://service.futuremark.com/compare?3dm05=1280176

No they are not the same. This thread is about PCMark05. The benchmark you ran was 3DMark05. They are different benchmarks.

[eva2000]

Raptors support TCQ, whereas SATA II drives support NCQ (different technology, you need a special controller for TCQ unless I am mistaken).

NF4 controller supports TCQ and NCQ http://nvidia.custhelp.com/cgi-bin/n...i=&p_topview=1

Question
Does nForce4 support command queuing?

Answer
nForce4 Ultra and nForce4 SLi can support tagged command queuing and native command queuing when used with SATA hard disks that support these features.

[IluvIntel]

My latest with rig in signature:

Note - user is aka "RodneyJM" and "Rojama" in other overclocking forums

7315

[Vapor]

You, sir, can get a LOT more from the 3D tests and your X800XT.

How are you getting such high other scores from your 630 though??

Especially your memory, it should be lower than an A64's by a noticable amount.

If you would, could you run a Sandra bandwidth?

[TuKo]

Is the Sillicon Image faster than nVidia's for RAID0 setup ?

[IluvIntel]

You, sir, can get a LOT more from the 3D tests and your X800XT.

How are you getting such high other scores from your 630 though??

Especially your memory, it should be lower than an A64's by a noticable amount.

If you would, could you run a Sandra bandwidth?

True, I could get more from the 3D tests, but don't want to at the moment with stock cooler on X800XT (waiting on new AC silencer 4 after bumping it and damaging a bracket... )

I did the benchmark with latest patch for PCMarkO5, creamed my old score of approx 5500. Seems pushing up the FSB with new bios for P5WD2 mobo really blows up the score...

Good thing about all of this as far as I'm concerned is that its all done with air/heatpipe cooling inside case and regular day-to-day background progames running.

Prescott power in 6xx series punches hard with right hardware support and not at great clock speed either...

[Vapor]

There's a certain point on the Asus boards where past that FSB you get very wrong scores (sends PLL timer out of whack, IIRC). Please run Sandra, I'd like to see if that's the case with you because there's no way your 6xx should beat a 3.1+GHz X2 in neither the multithreaded nor the memory tests.

Also, if you inch down your FSB, is there an inordinantly big jump at any one point?