• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

What do you use to bechmark your hard drive?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

DragonTattooz

Member
Joined
Aug 29, 2005
My Raptors don't seem to be running as fast as I thought they would. I base this on how long it takes Windows to load. How can I test the speed of the drives?

BTW- They are running in RAID 0, not sure of the stripe size. I just went with whatever the default was when I setup the array.
 

Tyreal

Member
Joined
Jun 5, 2005
Location
Vancouver, Canada
One program you can use is HDTach but remember that these benchmarks are all synthetic drives and do not directly symbolize real world performance.

I'll post more benchs for you in a sec.
 
OP
D

DragonTattooz

Member
Joined
Aug 29, 2005
Tyreal said:
One program you can use is HDTach but remember that these benchmarks are all synthetic drives and do not directly symbolize real world performance.

I'll post more benchs for you in a sec.

Thanks.

I'm using NVidia SATA ports 1&2 on my DFI mobo. They are supposed to be 3Mbps (I think) as opposed to 1.5 Mbps. Will the benchmark verify the speed of those ports?
 

jiggamanjb

Member
Joined
Dec 18, 2003
Location
MI, USA
I just downloaded that HDTach here is what I got for my single Raptor 74GB.


ics76o.jpg


Wonder what that fall off is at the 44GB range?????
 

Careface

Member
Joined
Jul 4, 2004
Location
Auckland, New Zealand
i dont know, but those results dont seem right.. my ordinary SATA (100mbps, not 150) barracuda gets 133.9MB/sec in burst speed, starts off at 60MB/sec and steadily drops off to 30MB/sec when its at the 160GB range.. i dunno, i guess i just assume your raptor should totally own my SATA (by a bigger margin that what you already have).. though i could just be expecting too much from it ;)

as for the dip in the 44gb range, it happens randomly. redo the test and i bet you it wont be there (if it is repeatedly there, perhaps there is something wrong with the drive at that point? someone wanna clarify that?)

EDIT: i just reread the post, raptors in RAID? theres definately something wrong there, it should be doing way better

Careface*
 

jiggamanjb

Member
Joined
Dec 18, 2003
Location
MI, USA
thegreek said:
defrag the drive

yea, i haven't done that yet. I just did a clean XP install and installed WoW, and that's about it. I'll do a defrag and see what happens. I still love this drive though :clap:
 
OP
D

DragonTattooz

Member
Joined
Aug 29, 2005
I'll try to upload a screenshot of my test. If this doesn't work, my results say 200 Mbps. Is that OK for 2 Raptors in RAID 0?
 

Attachments

  • Doc2.gif
    Doc2.gif
    42.5 KB · Views: 101
OP
D

DragonTattooz

Member
Joined
Aug 29, 2005
I suck at teh Interweb. :-/ How do I make it bigger without it becoming distorted? I have Acrobat 7 and Photoshop/Image Ready 7, but I suck at using them for anything but basic stuff.
 
OP
D

DragonTattooz

Member
Joined
Aug 29, 2005
Let's try this again. I think I just learned something...
 

Attachments

  • Doc2.gif
    Doc2.gif
    84 KB · Views: 205

GrayT

New Member
Joined
Dec 19, 2004
Are you using the 8mb zones? You should be using the 32mb zones to rule out those spikes. I noticed my speeds were a little higher with the 32mb zone.
 
Last edited:
OP
D

DragonTattooz

Member
Joined
Aug 29, 2005
GrayT said:
Are you using the 8mb zones? You should be using the 32mb zones to rule out those spikes. I noticed my speeds were a little higher with the 32mb zone.
I'm using the 32Mb zones. I've run the test several times and I always get those spikes. I also tried changing to SATA ports 3&4, but it didn't really make much difference.
 

GrayT

New Member
Joined
Dec 19, 2004
DragonTattooz said:
I'm using the 32Mb zones. I've run the test several times and I always get those spikes. I also tried changing to SATA ports 3&4, but it didn't really make much difference.

Hmm, well how is the real world performance of this setup? If it feels fine then there should be no reason to judge peformance off of these stats.

Still, it is very weird how one drive is getting such low scores. My 80GB SATA has burst speeds at 133mbps and it's a 7200rpm drive.

Also, I wouldn't worry about judging your speeds against the comparison drives that it, by default, displays. Those conditions are absolutely perfect lab testing benchmarks and I'm guessing are only there to theoretically show what is possible out of a certain drive type.
 

jiggamanjb

Member
Joined
Dec 18, 2003
Location
MI, USA
just grabbed the new drivers for my SATA controller and everything is working great now. just a little tip for anyone else having the same issue.
 

K15

Member
Joined
Nov 18, 2005
My 160GB Seagate 7200.9 starts at 75MB/s and drops to 40MB/s with an average of 63-65MB/s. The drop off is usually because another program accessed the drive while the test was running. Run it several times to get a good average. Your random access is amazing though, mine is 13ms, giving me a 8.8 seek :cool:
 

JCLW

Member
Joined
Apr 1, 2002
Ignore the burst speeds - they depend more on the controller then the drive. Only worry if your burst speed is as slow as your STR (or if your STR is as fast as your burst ;))

The two things you want to look at are the STR, and the access time.

jiggamanjb's graph looks fine for a single Raptor, ~70MB/s falling to ~50MB/s with an ~8ms access time.

DragonTattooz's graph looks wierd to me. The STR looks a little low, the graph is too flat (indicating a bottleneck somewhere around the 106MB/s mark), and the spikes go up instead of down. Access time is OK 'tho.

It doesn't matter if your SATA ports support SATA-300, because the Raptors only support UDMA-6 (ATA-133).

DragonTattooz: Try Irfanview (free) for quick easy image rotating & resizing. Try ImageShack for image hosting.
 

JCLW

Member
Joined
Apr 1, 2002
You can also try http://nodesoft.com/DiskBench/ (you'll need .NET)

The way I do it:
- Defrag
- Restart
- Kill things like A/V
- Load DiskBench
- Create the file "C:\bench.tst", block size 8MB and 256 blocks = 2GB, do NOT kill destination file
- Copy "C:\bench.tst" to "C:\bench2.tst", kill the destination file.
- Read "C:\bench.tst" in 32MB blocks
- Look at results, close DiskBench
- Delete C:\bench.tst