[Jim in Co]

Attached is a copy of my disk benchmarks. They seem low to me but I'm not sure what kind of numbers I should be looking for.

My HD is a WD Caviar Black, 500 GB, SATA, 3 GB/s with 32 MB cache.

Thanks in advance for any suggestions.

Jim

Attached File(s)

•  Test_Suite.png (26.74K)
  Number of downloads: 3

[Fiery]

Disk read results are definiately lower than expected. We tested the very same HDD, and we got the following results, in the same order than on your screen shot:

115.4 MB/s
99.3 MB/s
58.2 MB/s
82.2 MB/s
217.1 MB/s
12.07 ms

I'm not sure why there's a difference in read speeds. Make sure to check the SMART disk health status of your drive. Also make sure to close all possible background processes (e.g. Skype, MSN, Outlook, Media Player, all browser windows, etc) before running any benchmarks, since background processes could effect the benchmark results.


Regards,
Fiery

[athlonite]

this isn't so much as a drive issue but a combination of drive firmware and controller
as i have two Samsung HD502IJ (500GB) drives both on the same controller but both give very different results

the only difference is the port on the SATA controller

[Fiery]

Or maybe Samsung makes different hard disk drives with the same model ID It happened several times with WDC for example...

[jovan_YU1BCD]

Disk Drive #0 is 2 seagate 410as raid 0 on ICH10
Disk Drive #1 is 3 WD320AAKS raid 0 on ICH10
Disk Drive #2 is single WD640AAKS on Sil 3132 PCIe controler

I am little confused with buffer read marked with red on picture.

Attached File(s)

•  Test_Suit_745.jpg (238.81K)
  Number of downloads: 0

[IceMan]

here my report with 2 Samsung HDDs F1 Raid0

[Fiery]

Disk Drive #0 is 2 seagate 410as raid 0 on ICH10
Disk Drive #1 is 3 WD320AAKS raid 0 on ICH10
Disk Drive #2 is single WD640AAKS on Sil 3132 PCIe controler

I am little confused with buffer read marked with red on picture.

Buffered Read means the maximum bandwidth between the HDD (or RAID array) and the disk controller. It is the theoretical maximum read performance you can achieve with the current disk controller. In the case the Buffered Read score is very close to the Linear Read score, then there's a chance the interface (HDD/array <--> disk controller) is a bottleneck, so by moving to a higher bandwidth interface the disks could push out even more performance. But in your case the interface is not a problem

[jovan_YU1BCD]

Hi, IceMan

Do you believe that Everest help me to find spyware even some other antivirus and spyware programs
couldn't find?
From your results i see 'Warning situation', your hard drivers are ready to fail. Pay attention on "Linear Read (End)".
And they are very hot for sure. To avoid this issue just use technique "Short Stroking" (cut capacity for 15%).

This post has been edited by jovan_YU1BCD: 06 August 2009 - 07:48 PM

[athlonite]

well i have three samsung HD502IJ 500GB 16MB cache HDD's and two WD 250GB 8MB cache HDD's in raid0
the samsungs are all on the ATI SB600 sata controller whilst the WD's are on a Marvell 88SE6121 Serial ATA II Host Controller

now all the samsung drives whilst the basically the same are on 3 different ports on the SB600 they all give very different results as you can see from the pic

 HD502IJ_read_test_suite.png (90.62K)
Number of downloads: 0

as you can see very different test results which i've put down to slightly differing firmwares
Drive 0 1AA01110
Drive 1 1AA01113
Drive 2 1AA01113 (repaired controlller board)

[790]

This is test of 2006 year.
RAID-0 2xSeagate 7200.7, SATA-I, NCQ
ICH6R SATA-I (150)

Green Line: Drive BackCache is OFF
Red Line: Drive BackCache is ON

 DriveBackCache_OFF-ON.png (9.8K)
Number of downloads: 0
-----------------------------------------------
This is test of 11.2009
RAID-0 2xSeagate 7200.7, SATA-I, NCQ
ICH10R SATA-II (300)

 Drive400FSB-RAID-7200@7-CDEFG.png (5.96K)
Number of downloads: 0

5427MB/s is it better than SATA-III?

This post has been edited by 790: 02 November 2009 - 03:07 PM

[Fiery]

Obviously, the unusually high Buffered Read results are due to the activity of the BackCache. When you use any sort of caching (either software-level, Windows-level, hardware-level), such results could come up.