No announcement yet.

Disk performance Questions

  • Filter
  • Time
  • Show
Clear All
new posts

  • Disk performance Questions

    I am stumped. I am benchmarking one of our products and the disk performance marks are low. I am comparing IDE vs. SATA. I have verified that all necessary drivers are installed and IDE always beats SATA when run in a single drive configuration. If I boot the IDE drive and the use the SATA drive as a second drive and I benchmark the disk it blows the doors off of the IDE drive.
    This to me says that the SATA implementation on the board is probably not good, it appears that it is emulating IDE since it shows as an IDE drive, even in Windows (XP). This would mean some overhead in emulation and then the performance drop, is this possible?

    My question is, is the fact that the OS is running on the drive having that much of an impact? Should I load it up with memory and turn off the swap file to get a realistic throughput benchmark? I just don't understand why there would be a 5 fold increase in performance on a secondary drive over that same drive being used as a single drive.

  • #2
    What are the actual benchmark numbers and tests you are running. Maybe you are seeing cached RAM results rather than disk results.

    Older SATA hard drives contained a SATA to IDE bridge chip in them. But the new one are straight SATA.

    No the presence of the O/S on the drive should effect the results much unless there is some background disk activity going on.