Hi i want to buy an hd and i cant find any new on the read and write speed on a particlar hard disk

Discussion in 'Asus' started by DDC, Sep 23, 2005.

  1. DDC

    DDC Guest

    Hi i want to buy an hd and i cant find any news on the read and write
    speed on a particular hard disk drive like western digital DH


    The one that i want to buy is a Western Digital 160Gb S-ATA2 [7200]
    *8Mb*

    I'm switching from a ide uata133 setup to an sata2 hd... it will be
    fit with a
    Asus A8NE-FM (3PCI/PCI-E) NF4 Ultra [S939]
    or
    Asus A8N-E (3PCI/2PCI-E) NF4 Ultra [S939]

    Both are sata2 capable.
    And what i want to know is if there will be a difference
    from a ata 133 maxtor drive 80gig 8mb of cache.
    That ide drive run at 45 to 50 megs/s read and write on hdtach test so
    will i see some improvement with the new drive

    i hope to get around a conservative 70-90mb/s of data transfer as the
    drive can have some pick at 320mb/s.

    Is this ok?

    And also is there any difference between the two board here
    Asus A8NE-FM (3PCI/PCI-E) NF4 Ultra [S939]
    or
    Asus A8N-E (3PCI/2PCI-E) NF4 Ultra [S939]

    i think the first one is a sort of mini mobo and the other is a
    regular board. The problem is that i can't find anything one the Asus
    A8NE-FM (3PCI/PCI-E) NF4 Ultra. Is there something special???
     
    DDC, Sep 23, 2005
    #1
    1. Advertisements

  2. DDC

    chris Guest

    45 to 50 megs/s on HDTACH for a Maxtor 133/80/8 HDD....that's slow.
    Something is set wrong in whatever OS you are using.

    320mb/s = 40MB/s...so, no difference in the new drive.

    I don't know what OS you are using, but I'm getting 58-60MB/s with my
    100/80/2 Maxtor drive in Windows98SE.....
     
    chris, Sep 24, 2005
    #2
    1. Advertisements

  3. DDC

    Mercury Guest

    if you want HDD performance get raptors or go SCSI, then make sure the rest
    of your system has the performance to benefit. On top of that there is no
    point in getting max performance if the app s/w being used is inefficient.

    having SATA II over SATA I will give no benefit IMHO at this point in time.
     
    Mercury, Sep 24, 2005
    #3
  4. DDC

    DDC Guest

    My hard drive is a maxtor diamondmax plus 9 8mb 6Y080P0 uata133
    I'm using winxp pro sp2 and the hard drive is partitioned in 3.
    1 for the system the 2 for data and the last one is set for swapping.

    i'm using the nf2 5.1 driver. Witch one you use? Is your drive has
    only one partition.
     
    DDC, Sep 24, 2005
    #4
  5. DDC

    Paul Guest

    The A8NE-FM is rather interesting. It is a microATX board 9.6"x9.6".
    It doesn't have any PCI Express x1 slots, and has three PCI slots
    instead. It also offers Firewire, which I don't think the ordinary
    A8N-E has. The full title is "A8NE-FM/BULK", which means they
    are likely packaged 10 to a box, and the seller on this web page
    is breaking up the contents of the box for individual sales. (These
    boards would be used for OEM computer builds.)

    http://cgi.ebay.com.cn/ASUS-A8NE-FM...categoryZ93695QQcmdZViewItem#ebayphotohosting

    Even though I hate microATX motherboard, because usually the BIOS
    is missing features, at least in terms of the layout of the
    board, there aren't any useless x1 PCI Express slots on the
    board. (Now, the hard part will be finding a downloadable
    manual for this motherboard - without that manual, there
    is no way to tell in advance, what features are missing
    from the BIOS. MicroATX boards tend to have simplified
    BIOS, and I would read the PDF manual before purchasing such
    a motherboard. I cannot find A8NE-FM on the support web page!)

    As for your quest to find blistering speed on a disk drive,
    forget it! The Raptor is the fastest SATA disk on the
    storagereview.com performance database, and it is 71.8MB/sec
    at the beginning of the disk, and 53.9MB/sec near the end
    of the disk. The Raptor is a 10K RPM disk, which is faster
    than your average 7200 RPM SATA or IDE disk. It has superior
    seek speed, which speeds up compiles, makes the "find"
    command run faster, as well as having a good sustained
    transfer rate. Any time you deal with a large number of small
    files, a fast seek rate helps.

    The best 7200 RPM disk in the storagereview list, is the
    Seagate 7200.8, which is 69.8MB/sec at the beginning of the disk,
    and only 39.9MB/sec at the end of the disk.

    http://forums.storagereview.net/index.php?act=ST&f=2&t=18624

    To see transfer rates on some disks, select the two "WB99"
    entries in the menu on this page, then click "Sort" to see
    the information available.

    http://www.storagereview.com/comparison.html

    Also, when I go to the Seagate web site, and select >150GB
    capacity using the menu in the upper right corner here:

    http://www.seagate.com/products/interface/sata

    all the disks listed are 1.5gbit/sec interfaces. For example:

    ST3200826AS  Barracuda 7200.8 SATA NCQ  200 GB  SATA 1.5Gb/s 
    7200 RPM  8 ms avg  DPC=Desktop_PC

    I think that means that 3.0gbit/sec is not as important as
    you think :) Seagate is being deliberately evasive when
    discussing the issue of support for 3.0gbit/sec, when in fact
    it really would not help appreciably in the "real world". Right
    now, a 3.0gbit/sec disk interface, is just for bragging rights.

    On a completely different topic, I saw this item on using cloning
    tools with the NVraid. Read this before setting up a RAID 0 array
    on your A8N-SLI family board...

    (Cloning tool problems with NVraid - near bottom of page. See
    Mike Ramsden Mar 06, 2005, 10:39 PM posting...)
    http://www.hardwareanalysis.com/content/topic/38623?o=1120

    HTH,
    Paul
     
    Paul, Sep 24, 2005
    #5
  6. DDC

    chris Guest

    My hard drive is a maxtor diamondmax plus 9 8mb 6Y080P0 uata133
    I don't know much about Windows XP since I haven't worked with it much.
    What I can tell you is what I do in Windows 98.

    First of all, the swapfile size in Windows should always be 0. If it is
    greater then 0, add more memory! The reason for this is because using the
    swapfile has a drastic effect on Windows performance! I can not stress this
    point enough!!!

    As for having the swapfile on a seperate partition on the same drive, I
    don't see any benifit....and perhaps a slight slowdown of the system.
    Before defraging, disable the swapfile. Once defrag is finished, reenable
    the swapfile.

    As for drivers in Windows98, I don't use any special driver....again, check
    with someone in an XP NG for pointers. Both of my 80GB drives have 2
    partitions each. One partition for the OS and the other for the files I
    create.
     
    chris, Sep 24, 2005
    #6
  7. DDC

    Egil Solberg Guest

    Are you suggesting running windows without swapfile/pagefile? That is
    certainly wrong. It's best to let Windows handle this on its own.
    Before you tell me to back this up with examples etc, I suggest you do some
    research on your own finding articles on Google. Any credible source will
    tell you to stay away from making changes. If you aim for 0% pagefile usage,
    you're wasting lot of ram.

    In fact Microsoft says to avoid this. Again, let windows handle this on its
    own.


    ..
     
    Egil Solberg, Sep 25, 2005
    #7
  8. DDC

    chris Guest

    First of all, the swapfile size in Windows should always be 0. If it
    No, I'm not suggesting running Windows without a swapfile. Windows should
    be set to manage the swapfile on it's own. What I am saying is that the
    more often Windows uses the swapfile, the slower your system will be. There
    is a drastic difference in performance between swapfile and RAM usage. The
    more memory you put into your system, the lower the swapfile usage will
    be..hence, increased performance!

    The only time the swapfile should be disabled, is while running a defrag on
    the HDD. Once defraging is finished, you need to reenable the swapfile.
    This helps in getting rid of swapfile fragmentation.
     
    chris, Sep 25, 2005
    #8
  9. DDC

    Mercury Guest

    You choose to ignore that there is a break even point where the benefit of
    more memory doesn't justify the cost of the extra memory. Making
    generalisations about performance and memory is not beneficial as every
    system has different usage patterns that utilise resources uniquely - RAM is
    only 1 type of system resource.

    If a person needs more performance when that break even point is hit, they
    should invest funds elsewhere. If performance is critical, then they should
    have started with a sizing exercise where for example the outcome of the
    sizing may indicate that disc subsystem performance is more important.

    Memory is generally so cheap these days that it is seldom the cause of real
    performance bottlenecks for long because finally resellers are selling
    systems with an adequate minimum RAM and anyone that needs better
    performance will hopefully get some good advice quicky. People do add RAM to
    servers and desktops / laptops with less hesitation - due to the cost. (ECC
    RAM for example could almost be described as cheap).

    Adding more RAM is generally easy, but has its pitfalls in that command
    rates can drop from 1T to 2T and for performance critical apps, adding RAM
    willy nilly can be detrimental to performance.

    Having the swap file on its own partition can often result in longer seeks &
    guarantee them. This in turn reduces performance.
     
    Mercury, Sep 25, 2005
    #9
  10. DDC

    chris Guest

    Having the swap file on its own partition can often result in longer seeks
    &
    If the size of the swapfile is 0bytes, it doesn't matter which partition
    it's on. If the swapfile is greater then 0bytes and is on a partition on
    the same drive, I agree that this reduces performance. However, if that
    partition is on another drive ( other then the one the OS is on ), then
    there will be a gain in performance! Mind you, I have never tested the last
    statement for level of performance.
     
    chris, Sep 25, 2005
    #10
  11. DDC

    Mercury Guest

    Your argument is moot and prone to error.

    As soon as an app wants more ram than is available, you have no real VM
    without a swapfile so you will start getting odd errors.

    The errors can be almightily odd as one extremely common bad practice is
    apparently for programmers not to check that their requests for more ram
    actually worked. The consequences are so many and varied that the observed
    manifestation is entirely up to the app running and the part of the app that
    is active. One moment you may get a fatal exception, another, a menu option
    may not work, then next, results of calculations can be wrong and so on...

    It is better not to temp fate & Always have a swapfile.
     
    Mercury, Sep 25, 2005
    #11
  12. DDC

    chris Guest

    Ok. It's been a long day and maybe I'm not explaining myself correctly....

    I'm not telling anyone to set the swapfile to 0bytes. The swapfile on both
    of my systems ( running Win98SE ) is 0 bytes because I have enough RAM so
    that the swapfile is never used! I'm letting Windows handle the swapfile
    size ( as is the default setting ).

    Is that clear enough for you?
     
    chris, Sep 25, 2005
    #12
  13. DDC

    DDC Guest

    Hi, like you've said there not much performance gain in managing the
    swap file. And the only result that have shown some improvement is
    when i've set the paging file to zero. I don't know about the last
    statement that having the page file in a second hard drive will
    increase performance because my western digital is a lot more slow
    than my maxtor. Any way i've setup some result for you this way i will
    clear some thought


    My fist test on HDD benching.

    My setup is this one:
    a7n8x-x nf2 400
    driver: nf2 5.1 inter
    xp 2600 @ 1.916gig Barton 333
    1 gig of ram of 400 infineon

    My hdd is a 80gigs 2 recording surfaces maxtor diamondmax plus 9 8mb
    of cache

    So my first arrangement is this one.
    it has 3 partions, one is a 5 gig for the swap file. the second is for
    the os system and the last one is for data. This way the swap file is
    never fragmented. finally the swap file is on the same disk.

    With pcmark04 the hdd score is 4319
    xp start up : 8.155MB/s
    application loading : 6.144MB/s
    File copying: 32.25Mb/s
    General HDD usage : 5.026MB/s

    And the second test is : the swap file on the same partion of the os
    system

    With pcmark04 the hdd score is 4375
    xp start up : 8.313MB/s
    application loading : 6.309MB/s
    File copying: 32.226Mb/s
    General HDD usage : 5.054MB/s

    My third test is without any swap file (0 mb on hdd).

    With pcmark04 the hdd score is 4403
    xp start up : 8.353MB/s
    application loading : 6.339MB/s
    File copying: 32.715Mb/s
    General HDD usage : 5.071MB/s

    And my last test was made with a swap file placed on my second hard
    disk drive a western digital 40 gig 2mb of cache

    With pcmark04 the hdd score is 4334
    xp start up : 8.313MB/s
    application loading : 6.087MB/s
    File copying: 32.573Mb/s
    General HDD usage : 5.014MB/s


    Finaly the best way to get some performance out of your HHD is to set
    the paging file to 0 if you have at least 1gig of ram.
     
    DDC, Sep 25, 2005
    #13
  14. DDC

    deus maximus Guest

    You mean to say swapfile usage not size. Another correction to
    your suggestion concerns defragging. The swapfile should be
    defragged and kept that way: one eliminates fragmentation then
    fixes the minimum and maximum size at the same value and above
    that required for normal usage. This way the OS starts with a file
    that is not expanded by usage--no fragmentation-and large enough
    to run all apps. When I used win98, I set the swap to 250MB max
    and min after defragging the entire drive and never had to touch
    it again.
     
    deus maximus, Sep 25, 2005
    #14
  15. Windows 98 worked that way, I believe, but 2000 and XP always create the
    swapfile size they want. The default swap file size is never zero.
     
    Robert Hancock, Sep 25, 2005
    #15
  16. DDC

    chris Guest

    My hdd is a 80gigs 2 recording surfaces maxtor diamondmax plus 9 8mb
    NO!!! NO!!! NO!!! You do NOT set the swapfile to 0!!!

    As for your last test which moved the swapfile to your second HDD, that
    result is most likely flawed. To determine the real results, you'll have to
    use another drive which is the same model as your main drive.
     
    chris, Sep 25, 2005
    #16
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.