IOMeter v2006.07.27 – IOps

IOMeter v2006.07.27 


We continued our IOMeter testing with some (simulated) real world action – our standard PCPer quad of IOMeter profiles:

DDRdrive hits the ground running - PCI-E RAM-based SSD - Storage  1

DDRdrive hits the ground running - PCI-E RAM-based SSD - Storage  2

DDRdrive hits the ground running - PCI-E RAM-based SSD - Storage  3

DDRdrive hits the ground running - PCI-E RAM-based SSD - Storage  4

In our standard PCPer IOMeter testing, we see a very close race between the ioDrive and DDRdrive.  The extremely low latency enables the DDRdrive to reach top speed at very low queue depths.  The ioDrive is able to brute force its way past with greater bus bandwidth, but requires higher queue depths to accomplish the task.  The X25-M did manage to take a jab at the big boys, but only in the web server test, and only at higher queue depths.  We will be adding X25-E figures to the mix as soon as we get one in the lab.  The poor VelociRaptor remained plastered to the floor throughout all testing.

Note: For those thinking an I-RAM comes anywhere close to these numbers, I invite you to view this page for comparison. 


« PreviousNext »