The Notebook Review forums were hosted by TechTarget, who shut down them down on January 31, 2022. This static read-only archive was pulled by NBR forum users between January 20 and January 31, 2022, in an effort to make sure that the valuable technical information that had been posted on the forums is preserved. For current discussions, many NBR forum users moved over to NotebookTalk.net after the shutdown.
Problems? See this thread at archive.org.

    Disappointing RAID expresscard performance

    Discussion in 'Hardware Components and Aftermarket Upgrades' started by ruprecht_75, Oct 25, 2007.

  1. ruprecht_75

    ruprecht_75 Newbie

    Reputations:
    0
    Messages:
    3
    Likes Received:
    0
    Trophy Points:
    5
    Hi all,

    I've just got myself an Asus c90 and in order to get the most out of it I've decided to install external hard disks in a RAID 0 configuration. In order to do this I bought two 320Gb Seagate hard disks with 16Mb of cache each. I put them into an external enclosure that supports eSata. I then bought an eSata express card that supports RAID. I hooked it all up and managed to get it all going. However, I got some rather disappointing results and was wondering if anyone could offer an explanation as to why this is happening.

    Here's a screen dump from HD Tune of the external RAID disks:
    [​IMG]

    I also use a desktop at work which has RAID setup on it with similar hard disks. Here's the results for comparison:
    [​IMG]

    As you can see, the external RAID is about half as slow as the the setup on my desktop. From the graph, it looks as though there is some bottle neck that's stopping the hard disks from working to their potential. I'm guessing this is either from the actual express card or the express card interface. From my understanding, the express card interface can handle up to 2.5Gb/s which would seem more than enough for RAID. The other possibility is it's my expresscard. From what I've read expresscard slots implement two interfaces: PCI Express and USB2.0. Maybe the card is using the expresscard's usb interface?
    ( http://www.expresscard.org/web/site/about.jsp )


    Does anyone have any ideas?
     
  2. Greg

    Greg Notebook Nobel Laureate

    Reputations:
    7,857
    Messages:
    16,212
    Likes Received:
    58
    Trophy Points:
    466
    The problem is the card. The IC in there can only work so fast!

    The USB interface would top out at 40-60MBps, and you're seeing around 90MB/s so I know you aren't using that bus.
     
  3. tebore

    tebore Notebook Evangelist

    Reputations:
    55
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    30
    That card can only work so fast and as you can see there's quite the CPU overhead. The card doesn't have a dedicated CPU and is using your CPU to help offload and that process has a lot of latency and overhead.

    You're gonna need a really good card to get decent performance out of these high density drives nowadays.
     
  4. ruprecht_75

    ruprecht_75 Newbie

    Reputations:
    0
    Messages:
    3
    Likes Received:
    0
    Trophy Points:
    5
    I thought it might have been the card. I did get it off ebay.

    Would it be worth investing in a better card? Say the SIIG eSATA II ExpressCard for example?
    http://www.siig.com/ViewProduct.aspx?pn=SC-SAE612-S1

    Or will this have the same problem? Are there any cards out there that can handle the full speed of these drives?
     
  5. Greg

    Greg Notebook Nobel Laureate

    Reputations:
    7,857
    Messages:
    16,212
    Likes Received:
    58
    Trophy Points:
    466
    As far as I know, all these eSata cards have ICs built from the same manufacturer...so I kind of doubt getting a new one is going to help. That being said...those drives are still pretty darn fast!
     
  6. tebore

    tebore Notebook Evangelist

    Reputations:
    55
    Messages:
    521
    Likes Received:
    0
    Trophy Points:
    30
    He's pretty much getting single drive performance. I guess the IC in the card somehow can't manage working with both drives at the same time.
     
  7. ruprecht_75

    ruprecht_75 Newbie

    Reputations:
    0
    Messages:
    3
    Likes Received:
    0
    Trophy Points:
    5
    Ok, i've done some more testing, this time on a dell 6400. For reference here's the results for the RAID 0 config:

    [​IMG]

    I then set them up as individual drives. Here's the result of disk 1 (disk 2 is identical):

    [​IMG]

    I then ran hd tune on disk1 and disk2 simultaneously. here are the results:

    [​IMG]
    [​IMG]

    Each disk only performs half as well. It looks as though there is a limitation in the card, as though it can only handle a transfer rate of around 70MB/s.

    However, from the following review,
    http://forum.notebookreview.com/showthread.php?t=84710

    it appears that there are cards that can handle the speed of two disks.
     
  8. ViciousXUSMC

    ViciousXUSMC Master Viking NBR Reviewer

    Reputations:
    11,461
    Messages:
    16,824
    Likes Received:
    76
    Trophy Points:
    466
    As with any full bandwidth hdd test you can see the curve to the hdtune test. Yours is flat all the way across so you can clearly see there is a bottleneck. Be it the IC' of the card, or the interface its running thru. However Raid0 is severely over rated in the first place. In a benchmark you may see double the speed, but in real life use its never going to be that unless your transferring/saving/opening a very large file. Its good for servers and for video editors and things.

    As a gamer or everyday computer user tho the file sizes are mostly below the strip size and performance can actually be lower than a single drive.

    I say cut your losses and run a single disk, or use a different raid configuration or just keep it as is if you really want too because its not much slower than your setup from work when you take into account average transfer speeds.

    Plus does your setup at work have the same drives? They may be SCSI or 10,000 rpm drives ect.