Post your hard drive benchmarks!

[ih8modem]

fyi

Gigabyte P965 DS3 , 4GB RAM , Vista Home Premium , attached to same ICH8 port for all 3. In each case the drive is just attached, nothing on it at all. The boot drive is a WD2500
Samsung F1 1TB 3x334MB platters

WD6400AAKS 2 x 320GB platters

WD3200AAKS - 1 x 320GB platter

 

[slowrunner]

spinpoint 200, 50GB partion for c:, the rest for D:

think it could do with reverting back to one partition and using just as D:, with a faster drive for c:, What do you think?

 

[ih8modem]

CF UDMA USB2.0 reader

This is a Lexar 4GB 300x UDMA compactflash on a USB2.0 UDMA Lexar Pro reader.

 

[ih8modem]

Same Card IDE attached

Same 4GB card attached to an IDE/CF adapter (UDMA capable).

 

[Mint_Sauce]

Perc 5/i with LSI Firmware and now with 512mb RAM. 5x500gb WD RE2 in raid5.

 

[sprognak]

ICH9 running 4x Samsung 1Tb F1 drives, E2200 CPU, 4Gb Vista Gold OCZ (all at stock). Abit IP35, Windows 2003R2 64 bit.

 

[1kiwi]

F1 1tb



Raptor 74gb

 

[fobose]

Just got hold of a new drive for my laptop

I got one of these drives when they first came out (cost me £97inc on OcUK on "this week only"). I use mine as an external backup though.

 

[balls]

4 disk raid 0 using 7200.10 160gb disks

 

[BruceLee]

I got one of these drives when they first came out (cost me £97inc on OcUK on "this week only"). I use mine as an external backup though.

Got mine for £80, cant believe the prices of storage these days.

Swapped mine out with the Hitachi 160Gb in my laptop, seems a lot better, transfer speeds across the disk are over twice as fast and its damn near silent. Using the 160Gb in an external caddy.

 

[valentha]

2 x Samsung HD501LJ raid 0 first 120gig

2 x WD740GD Raptors raid 0

Now which shall I put the OS on

SB600 southbridge

 

[Simmy]

Do you actually see a big performance increase in real terms?

Compared to the drives running on their own, yes XP is a lot more responsive. Based on some of the benchmarks in this thread, it looks like I might have to sell my drives though and buy 2 samsung F1 drives instead

 

[1kiwi]

Still can't beat those access times. Why not have the best of both worlds.

 

[valentha]

Still can't beat those access times. Why not have the best of both worlds.

Well stuck the OS on the samsung small partition, swap file and IE cache onto the raptors which leaves an 880gig partition for dumping anything else on . it flies

 

[Netvyper]

2 Samsung 1tb F1s in RAID0: (Comparison against 6x 500gb Seagates)

4 Samsung 1tb F1s in RAID5:

I figure there must be something limiting me there, that isnt the disks. Does anyone have any suggestions

Edit: This is using a PCI-E 4x Highpoint 2320 Controller. With a s939 X2 3800+, 1gb DDR at 2700.
I am just trying write through cache rather than write back.

Edit2:
HD-Tune says Similar:

 

[rpstewart]

I had a similar problem with my RAID5 array - go into the HDTune settings (little cogs next to the exit button) and increase the block size in the Benchmark settings.

You don't want write through caching btw, writes will be slower than a paraplegic sloth.

 

[Netvyper]

I had a similar problem with my RAID5 array - go into the HDTune settings (little cogs next to the exit button) and increase the block size in the Benchmark settings.

You don't want write through caching btw, writes will be slower than a paraplegic sloth.

That seems to give it a decent sustained speed, but the burst is lower than the sustained
EDIT: is that cause the burst is too short for the disks to align and work together?

and yeah, the cache changes were just for testing, now firmly back into write back caching

8MB Block, Full Accuracy:

 

[rpstewart]

but the burst is lower than the sustained

Same here. I didn't bother digging too deep to find out why though.

 

[ih8modem]

ever used crystalmark

Has anyone ever use this benchmark before

http://crystalmark.info/?lang=en

Here is a screenshot of the bench in use :
Lexar 4GB 300x CF card

Samsung 1TB F1

 

[1kiwi]

why is the write faster than the read?