OK, so let's say I did get you real numbers from a synthetic benchmark. What does that tell you about what your experience will be with YOUR file system that will be filled differently from mine, with YOUR applications which are different from mine?
Synthetic benchmarks have their place (believe me, I am in Technical Marketing for an enterprise storage company) but they aren't the final answer.
Many people don't understand that throughput isn't the most important measure of hard drive performance. You need good throughput with good seek times. On top of that, these numbers by themselves don't mean diddly in the real world because laptops are doing many threads at once, even if you think you are using only one application. NCQ was designed to fix a deficiency in SATA when compared to FC and SCSI drives.
You can have 2 drives, one SCSI and one ATA (non NCQ), with similar throughput and seek times, and similar benchmarks using IOMETER. The SCSI drive, all things being equal, will do a better job when the system starts thrashing with multiple threads, multiple apps, etc. Heck, ATA and SATA drives stink when doing mixed read/write workloads, and a SCSI or FC drive will just eat them alive.
So don't go by XBENCH or IOMETER or other synthetic benchmarks as the end-all-be-all of driver performance. Get opinions from people you trust. If you don't trust anyone, get better friends (grin).
