Testbed and method
The benchmarks were performed on a system equipped with:
- PhenomII 940 CPU (4 cores @ 3.0 GHz, 1.8 GHz Northbridge and L3 cache)
- 8 GB of DDR2-800 DRAM (operating in unganged mode)
- Asus M4A78 Pro motherboard (AMD 780G + SB700 chipset)
- a single 500 GB 7200-RPM Seagate disk in AHCI mode
- S.O. Linux RHEL 6.3 x64
I tested with both NCQ and software queue ON/OFF. By default, the OS queue is 128 element long, but I also benchmarked with a 1024-element queue size.
I used a simple benchmark called "seeker" that overload the disk with random read requests. In this scenario, the DRAM disk cache should be ineffective to mask disk latency, so we should see some significant differences between testing modes.