Recent content by MikeU

  1. MikeU

    The Southbridge Battle: nforce 6 MCP vs. ICH7 vs. ICH8

    Yes, but with six SAS 15k drives running off an Adaptec 4805SAS PCIe 8x RAID controller, so the chipset isn't a big factor in disk performance. Unfortunately, I don't have multiple SATA drives to try in RAID 0 using the southbridge on a 975x motherboard and 680i motherboard to see what IOMeter...
  2. MikeU

    The Southbridge Battle: nforce 6 MCP vs. ICH7 vs. ICH8

    Have you found any testing using something other than HD Tach, which has a history of problems measuring performance on RAID arrays, comparing ICH7 or ICH8 RAID performance with 680i RAID performance? HD Tach shows an average read speed of about 300 MB/s for my Adaptec 4805SAS RAID array, with...
  3. MikeU

    The Southbridge Battle: nforce 6 MCP vs. ICH7 vs. ICH8

    This still appears to be valid if IOMeter is a valid benchmark for disk I/O. However, my theories about the reasons for the differences turn out not to be valid. It seems to be necessary to reboot before each IOMeter test in order to get consistent results, and changing the number of...
  4. MikeU

    The Southbridge Battle: nforce 6 MCP vs. ICH7 vs. ICH8

    The wall may be in HD Tach rather than the nVidia hardware or software. I ran into a similar bottleneck using HD Tach at about 300 MB/s with a RAID 0 array of six Hitachi Ultrastar 15k SAS drives on an Adaptec 4805SAS PCIe 8x controller. When I contacted Adaptec, I was advised to use IOMeter...
  5. MikeU

    The Southbridge Battle: nforce 6 MCP vs. ICH7 vs. ICH8

    How accurate is HDTach? I'm involved in a discussion about the southbridge RAID performance comparisons in the eVGA Motherboard forum. I mentioned that the HDTach read test showed a straight line at about 300 MB/s with no change in speed across an array of six Hitachi Ultrastar 15k 34-gig SAS...