WIN2000 RAID 0 Not Faster than 1 Drive on its own?

[james_is_great]

I have just installed WIN2K On one 20GB IBM60GXP and have 3 of the same type of drives set as RAID 0. All hard Drives are set as Masters and connected individualy to 4 IDE ports.

So why when i Benchmark the RAID 0 (3xDrives) and the Boot drive in Sisoft SANDRA 2001 PRO do they score very similar scores.
Results as follows of RAID 0 3 IBM60GXP 20GB all on seperate IDE channels.
Buffered Read 42MB/s
Sequential Read 48MB/s
Random Read 6MB/s
Buffered Write 25MB/s
Sequential Write 30MB/s
Random Write 10MB/s
Average Access 9ms
Score 29814
Don't make a lot of sense Sequential being faster that Buffered! Something is wrong but what?

Results for IBM60GXP 20GB Drive on its own.
Buffered Read 44MB/s
Sequential Read 38MB/s
Random Read 7MB/s
Buffered Write 26MB/s
Sequential Write 26MB/s
Random Write 8MB/s
Average Access 7ms
Score 25050
RAID 0 isn't much faster and sometimes slower.

I have an ABIT KT7A-RAID 1GHz 133FSB Athlon, 256MB CAS2 Mem, Sound Blaster Live PCI port 6, 3COM 10/100 NIC PCI port 3 And Geforce DDR.

I have Installed VIA 4in1 ver1433 and HPT370 ver612

I did originally have the all 4 drives on the RAID contrler but performance wasnt any better than what i have with 3 drives under WIN2k RAID 0.

Can anyone help? I feel that i have been duped into buying a so-called performance MB when really its just a pig in a poke. I had better RAID 0 under WIN2K Performance with my trusty BP6 with just 2 Drives.

Find a job you love and you will never work a day.

 

[CALV]

Personally I dont belive a word sandra says, for eg, when I benchmark my memory, (corsair pc150 cas2 currently running 140 fsb so ram is doing 140), sandra gives it a score that was about that of a 800 duron with pc100 ram (or similar, I cant remember exactly but compared it to something a lot less), so I shouldnt take sandra scores for granted.


Next time you wave - use all your fingers

 

[james_is_great]

CALY you might be right but I am comparing SANDRA scores with other hardware on my computer so if the RAID scores arnt reported corectly then surely the scores for my other hard drive on its own would have the same error.

Find a job you love and you will never work a day.

 

[CALV]

hmm, yeah I see your point, even if the scores were miles out, you should see a relative miles out increase!. Have you tried hdtach, its a very good hdd speed measuring test, the unregistered version can be had from <A HREF="http://www.calvsplace.cwc.net" target="_new">my site</A> If you need erm info on erm how to register, then mail me calv@fuckyou.co.uk


Next time you wave - use all your fingers

 

[james_is_great]

Thanx for the offer but i already have it erm with the Reg. I have also been using benchmark feature of Virtualdub 1.4 which is a program i use quite a lot. I am now testing the drives individualy under win2k to see if maybe one of the drives has a problem.
I took out sound and NIC card and got 15% boost then loaded safe settings in bios and Changed computer to standard PC instead of ACPI Now i get about 55MB/s Sequential Read and 78MB/s Buffered Read about 33% increase so far. Still testing so far and might have a few things to add to pauls unofficial KT7 FAQ


Find a job you love and you will never work a day.

 

[Guest]

I wouldn't think just 2 drives in a RAID 0 array under Win2k would stress a 1GHz machine, but that is what it is pointing toward as it's still software dependant to some degree. Wonder if it's a driver issue for that onboard controller.. might be that as well.

***check the jumpers 1st then check em again***