Uh-oh, iMac i7 shipped with faulty hd?

Discussion in 'iMac' started by porly, Jul 7, 2011.

  1. porly, Jul 7, 2011
    Last edited: Jul 7, 2011

    porly macrumors newbie

    Joined:
    Jul 7, 2011
    #1
    imac 2011 core i7 weird hd tune results via bootcamp win7 x64.
    this can't be right, can it? help! any hd tune results for this machine?
    edit: noobish hysteria on my part. 'tis all good.

    [​IMG]
     
  2. 88 King, Jul 7, 2011
    Last edited: Jul 7, 2011

    88 King macrumors 6502

    88 King

    Joined:
    Jun 18, 2011
    Location:
    London, UK
    #2
    That looks OK to me. Try Crystal Disk Mark in Windows, or xbench in OSX and report back as these programs test both read and write speed at the same time.

    EDIT: Here is my xbench scores of the same hard drive I did a week ago.
     

    Attached Files:

  3. porly thread starter macrumors newbie

    Joined:
    Jul 7, 2011
    #3
    with "turn off Windows write-cache buffer flushing" enabled i got
    a nice evened out result bar any drops. that previous graph just doesn't look right. i made a few hd tests in the past and they were never that extreme.
    plus with bootcamp while copying files the machine is barely usable.
     
  4. porly thread starter macrumors newbie

    Joined:
    Jul 7, 2011
    #4
    seems to be a problem with win7. got about the same results with xbench.
    [​IMG]
    thank you.
     
  5. 88 King, Jul 7, 2011
    Last edited: Jul 7, 2011

    88 King macrumors 6502

    88 King

    Joined:
    Jun 18, 2011
    Location:
    London, UK
    #5
    I can suggest those tools for diagnosis. You can always ask Apple for a hard drive replacement or a new machine. I know its a hassle as a new i7 machine will take a week to arrive.

    Diagnostic tools for Windows

    Run check disk utility

    Try Pass Mark Disk Checkup
    http://www.passmark.com/products/diskcheckup.htm

    SeaTools, Seagate’s dignosit tool
    http://www.seagate.com/ww/v/index.j...toid=720bd20cacdec010VgnVCM100000dd04090aRCRD

    MHDD, a low level hard disk diagnostic tool
    http://hddguru.com/software/2005.10.02-MHDD/

    You can also run Apple Hardware Test by insert OSX installation DVD and hold down D after reboot.

    EDIT: Never mind, start to write before your last post.
    Here is an idea. During Windows 7 installation, there was an option on which hard drive partition to install Windows, I deleted the Bootcamp part and reformatted it to NTFS, and its been working fine. Not sure if its worth your time to reinstall Windows again.
     
  6. MacHamster68, Jul 7, 2011
    Last edited: Jul 7, 2011

    MacHamster68 macrumors 68040

    MacHamster68

    Joined:
    Sep 17, 2009
    #6
    where is the problem? what makes you think a hdd would be faster in a i7 then in a core duo , the hdd has still the same speed s-ata and 7200rpm , so loading apps will make not much of a difference and moving files wont be faster either , you need a faster HDD or a ssd to see a difference, as the hdd is the bottleneck in that case
     
  7. velocityg4 macrumors 68040

    velocityg4

    Joined:
    Dec 19, 2004
    Location:
    Georgia
    #7
    I agree. In fact those results look pretty darn good for a hard drive. I know the last "fast" hard drive I bought peaked at about 70MB/s and that was the fastest 7,200RPM HD on the market. Now that I am transitioning boot drives to SSD and using HD's for bulk data storage speed is secondary.
     
  8. porly thread starter macrumors newbie

    Joined:
    Jul 7, 2011
    #8
    i did not expect it to be faster. it seemed all around sluggish to me.
    now the windows side is fixed, albeit risking data loss due to disabled flushing.
     
  9. chown33 macrumors 604

    Joined:
    Aug 9, 2009
    #9
    The only technical reason to flush the write cache is if you're expecting power loss at any moment, and you want to help ensure data integrity. An Uninterruptible Power Supply (UPS) helps ensure data integrity by mitigating power loss.

    I think everyone who cares about data loss should have an UPS anyway, regardless of whether they've enabled or disabled write-cache flushing.
     
  10. porly thread starter macrumors newbie

    Joined:
    Jul 7, 2011
    #10
    apparently you're quite knowledgeable about this.
    why would i not have to tamper with these settings in the past and
    still get a graph with less extreme amplitudes?
     
  11. chown33 macrumors 604

    Joined:
    Aug 9, 2009
    #11
    Maybe your previous hard drive didn't have a write cache. Or if it had one, maybe it wasn't as large.

    If the write cache is relatively small, then flushing it is less of a performance hit. Larger caches take more time to write to disk, and it's the writing-to-disk part that causes the slowdown. The write won't be completed until the whole cache is written to disk. Since the purpose of the cache is to avoid paying the penalty for writing to disk, any time the cache is flushed, you've slowed things down to the same speed as if there were no cache at all.

    It's like the difference between stop-and-go driving, where you can drive at 50 mph between traffic lights, but if you hit some red lights, you're going 0 mph a significant part of the time, which kills your average speed. It's possible that a steady 10 mph on a bicycle could beat stop-and-go driving in a car. Flushing the cache is like a stoplight: the computer won't proceed until the entire cache contents is written to disk. So if you have a bigger cache, the computer is stopping for longer periods of time.
     
  12. basher macrumors 6502

    basher

    Joined:
    May 27, 2011
    Location:
    Glendale, AZ USA
    #12
    Performance stats look good. What were you expecting?
     

Share This Page