How fast are your hard drives?

[Joepublic2]

I found a neat little program that's easy to use called Quickbench.
http://www.benchmarkhq.ru/fclick/fclick.php?fid=199
It's a fast little test that gives sequential read/write performance for a hard drive and random read performance with various block sizes. The default is to test up to 1MB block sizes. Here are my results:

PC 1: Nforce 3 250 chipset with Seagate 160GB ATA100 ST3160023A drive
http://img131.echo.cx/img131/5852/result16sl.png

PC 2: Intel 82443BX (440BX northbridge) with Maxtor 60GB ATA33 6L060J3 drive
http://img33.echo.cx/img33/9734/result28kt.png

How do your hard drives stack up?

Thanks to ImageShack for the free image hosting.

 

[Intelman07]

Here is mine. It is a seagate 200GB 7200.7 SATA drive.

http://www.imageshack.us][img=[url="http://img33.echo.cx/img33/3639/harddrive6hn.png"]http://img33.echo.cx/img33/3639/harddrive6hn.png[/url]][/url]

 

[mechBgon]

Different benchmark, but here's one of mine, sequential read speed & random-access : pic. Not bad for a "has-been" drive

(edit: Cheetah 15k.3 on LSI Logic U160 PCI card there)

 

[Joepublic2]

Here are my HD Tach numbers.

http://img218.echo.cx/img218/5488/result34kh.png

Are those big dips normal? I'm running XP64, and the Nforce drivers still seem a little bit immature to me. Could they cause those dips, or is it the drive itself?

 

[mechBgon]

My 250GB SATA 7200.8 on nF3 shows those types of dips too, moreso than yours. HDTach tends to do that for some reason; most of my Cheetah 15k.3 benchies also have at least a few dips. Naturally I picked the nice smooth one to post a pic of

From looking at the pic, you'd say "oh, he gets 75MB/sec, I get 60MB/sec, therefore his is 25% faster," but in seek-intensive work, the performance advantage of the Cheetah 15k.3 over the SATA 7200.8 is up to ~100%, presumably due to the very low seek times and (working) command queueing.

In some situations, you would notice the difference, in some you would not. So take all the benchies with a grain of salt, it's reality that counts in the end I like Anand's recent drive benchmarks that include game loading times, that's good reality-based benchmarks to look at.

 

[montag451]

Hi, I had a few minutes spare, so I thought I would test the following drives that I had around under varying conditions. THIS IS IN NO WAY a complete test, but, the results were pretty surprising nonetheless [to me anyway].

The system was as follows:
ABIT NF8
754pin AMD64 3400+ [512KB L2]/Stock heatsink+fan
Corsair Value 512MB DDR PC3200
PSU = TAGAN 480 [Recommend highly]

The drives tested were:
200GB Maxtor - 6Y200P0 (Primary MASTER)
160GB Seagate SATA - ST3160023AS (SATA 1)
20GB Maxtor - 6L020J1 (Secondary SLAVE)

They were tested with HDTach 2.61 under the following conditions:
1) Fresh install XP SP1 slipstreamed --- No chipset drivers installed
2) SP2 installed --- No chipset drivers installed
3) SP2 installed --- Chipset drivers installed

Each drive was tested 3 times under the Fresh install with SP1, then once after the SP2 was installed, then once more when the chipset drivers were installed.
Each drive was tested once, then then next drive, etc, then back to the first drive again.

.....................................TEST 1 RESULTS - SP1 - No chipset drivers
200GB MAXTOR - 6Y200P0
..Random Access Time [ms]...MAX read....MIN read....AVERAGE read....CPU utilisation
1).............15.3.............................65210......... 3644.............51865.................23.5%
2).............14.8.............................65204.........39288............51929.................20.4%
3).............15.2.............................65211.........39535............51932.................22.6%

160GB SEAGATE SATA - ST3160023AS
1).............12.7.............................57387.........17809............45851.................19.7%
2).............12.8.............................57126.........17778............45762.................20.2%
3).............12.8.............................56670.........14839............45851.................21.3%

20GB MAXTOR - 6L020J1
1).............12.5.............................43262.........21858............34487.................19.9%
2).............12.6.............................42316.........21861............34487.................20.1%
3).............12.5.............................42311.........21858............34484.................21.2%



.....................................TEST 2 RESULTS - SP2 -- NO chipset drivers installed
200GB MAXTOR - 6Y200P0
1)...............15.1.............................65225......... 4035............51898.................24.0%

160GB SEAGATE SATA - ST3160023AS
1)...............12.7.............................57594.........22692............45876.................22.4%

20GB MAXTOR - 6L020J1
1)..............12.7.............................42319.........21858............34486.................23.0%



.....................................TEST 3 RESULTS - SP2 -- Chipset drivers installed
200GB MAXTOR - 6Y200P0
1)..............15.7.............................65211......... 1188............51625.................20.7%

160GB SEAGATE SATA - ST3160023AS
1)..............12.7.............................57696........20880............45851.................22.0%

20GB MAXTOR - 6L020J1
1)..............12.7............................42315.........21859............35544.................22.2%






Well, there you have it. ALL the drives on all the tests maxed out at HDTach's Read Burst Speed @80MB.
No other software/drivers were installed apart from OS, and where specified, the chipset drivers.

 

[mwmorph]

wow. i thought my hdd was slow, being a oem samsung/hp unit and all running on a sis mATX mb using old UDMA. The drive is comparable to the op's first drive. nice. espically since it's set at quiet mode in AAM acoustic management.

 

[Joepublic2]

Originally posted by: mechBgon
From looking at the pic, you'd say "oh, he gets 75MB/sec, I get 60MB/sec, therefore his is 25% faster," but in seek-intensive work, the performance advantage of the Cheetah 15k.3 over the SATA 7200.8 is up to ~100%, presumably due to the very low seek times and (working) command queueing.

I know, I was mainly just doing this to make sure my hard drive wasn't suffering from some performance malady, as it seems to take twice as long to load levels in UT2004 in XP64 than it did in XP32. Since the throughput isn't off for a drive of it's class, I guess the drivers are using more CPU time, or it's an issue specific to UT2004.

Originally posted by: mwmorph
wow. i thought my hdd was slow, being a oem samsung/hp unit and all running on a sis mATX mb using old UDMA. The drive is comparable to the op's first drive. nice. espically since it's set at quiet mode in AAM acoustic management.

I don't know what chipset you're using, but I believe that SIS's later Athlon XP and all their Athlon 64 chipsets have had very good disk performance.

 

[Sforsyth]

Here I ran both HD TEch and that benchmark program on my brand new Maxtor 250 GB 16 MB Cache hard drive that I got last Friday.

http://home.cogeco.ca/~scottforsyth/HD_Bench.JPG

http://home.cogeco.ca/~scottforsyth/HD_Tech.JPG

do these seem right ?

 

[Pariah]

Originally posted by: mechBgon
Different benchmark, but here's one of mine, sequential read speed & random-access : pic. Not bad for a "has-been" drive

(edit: Cheetah 15k.3 on LSI Logic U160 PCI card there)

Very funny mech. The 15k.3 is not a "has-been" drive. It's so good, in fact, Seagate decided to re-release it almost 2 years later with a capacity bump as the 15k.4.

 

[Chaotic42]

37.26 MB/s

Seagate 10K 18.3GB SCSI

 

[Thunderstick]

http://img236.echo.cx/my.php?image=hdd1kx.jpg]

[/URL]

 

[Thunderstick]

54.57mb read
46.78mb write

seagate 200gb ATA/100 7200.7

 

[bob4432]

number seem a little funny compared to hdtach but here they are:

WD80GB 8MB Cache

IBM 60GB 8MB Cache

Fujitsu 10krpm 36GB 8MB Cache u320 on a LSI U160 Card

 

[stevty2889]

2x36GB Raptors Raid 0