Is Sandra's File System (HDD) benchmark reliable??

[LarryJoe]

I seem to get a drastically different score on my RAID 0 everytime I run the thing. Actually, it seemed to give me inconsistent scores on my old Promise Utlra100 controller as well. I tried to use HD Tach, but the demo will not run under XP as it requires a purchase to be made for NT environments. I have a problem shelling out $50 just to benchmark my HDD.

Thanks,

LJ

 

[ojai00]

I read in another post that Sandra's hard drive benchmarks aren't very accurate. It was suggested that you get another dedicated program, but I can't remember it off the top of my head. Hopefully someone else can stop by and provide that link. Hope this helps.

 

[MichaelD]

Hello there,

Back in my Win98 *snicker* days, I used HDTach and liked it a lot, but as you already found out, it doesn't run on XP/W2K.

When you say you get "drastically diff" results, what do you mean, exactly? A few MB off here and there is within the acceptable range. Maybe you have stuff going on in the background while your benchmarking? Disable NAV, IM, etc.

SANDRA isn't the best thing out there, but it's commonly used and accepted as a reasonably accurate benchmarking program..

 

[LarryJoe]

Thanks. When I first set up my RAID 0 a week ago and installed XP, I got a score of +3000, a week later and I'm at 1700. I don't have alot running in the background, but you never know. I probably did the 3000 before installing stuff. I have MBM, Volume (Santa Cruz CP), and Max10 (remote control software). That's it.

I'll kill these and give it another run tonight.

 

[Kwad Guy]

The last time I ran the Sandra HDD benchmark, it told me that
my Maxtor 7200rpm 45gb hard drive was a lot faster than my
2 x IBM 10K raid 0 array running off a AMI SCSI Ultra160 raid array
running in a 64bit 66mhz slot.

Don't know if I was doing something wrong, but that kind of turned
me off to the value of the Sandra benchmarks.

Kwad

 

[lorlabnew]

I had the exactly same problem (no HDTach on XP, inconsistent results in Sandra), and someone around here suggested WinBench '99 (with it's Disk WinMark tests); I like it much more than Sandra, seems to be more consistent in results, and definitely more sophisticated. Also, it's used by many reviewers of HW, so you can compare your results easily. And it's free.

 

[Agent004]

The best storage site, Storagereview, dislikes Sandra as it's too theortical and not that practical, doesn't tell you much about the real world usage of the drive.

 

[LarryJoe]

Thanks. I checked out storagereview. Nice site and found they also highly recommend Winbench 99. Great advice, thanks again. I'll give it a shot tonight.

 

[Rand]

Sandra is worthless as a HDD brenchmark.
ATTO is better then Sandra but still not very good.
HDTach is okay for a basic indicator of performance.
WinBench is excellent and very thorough, but is occasionally specifically optimized for which makes some results a touch dubious... still it's the best desktop HDD benchmarking program easily obtainable.

IOMeter is great for a server HDD benchmark, but very poor for desktop performance.
Intel's IPeak is extremely good, but also expensive and very hard to find.

 

[LarryJoe]

Thanks Rand. I just loaded up Winbench. Which Disk benchmark should I run? There are many to choose from.

 

[Rand]

<< Thanks Rand. I just loaded up Winbench. Which Disk benchmark should I run? There are many to choose from. >>



All of them to get the most thorough idea of the relative performance characteristics of your drive.

 

[LarryJoe]

I am getting an error message that says it can't overwrite a file that already exists. I have not been able to run this yet, I wonder if it is compatable with XP?