HD Tach is a low level hardware benchmark for random access
read/write storage devices such as hard drives, removable drives
(ZIP/JAZZ), flash devices, and RAID arrays. HD Tach uses custom device
drivers and other low level Windows interfaces to bypass as many layers
of software as possible and get as close to the physical performance of
the device possible.
SATA II 3.0GB/s Performance:
SATA III 6.0GB/s Performance:
Benchmark Results: Here in HD Tach we see that really the only performance difference is seen in the Burst Speed test, which went from 190.0 MB/s to 1853.2 MB/s! This is impressive, but we also see something that worries us in the write test when using the Marvell controller for SATA III 6.0 GB/s performance: the spike at the very beginning of the benchmark.
It was discovered that the Marvell driver allocates a portion of system memory to perform caching duties. This basically means that the drive by default has “Write caching” enabled. “Write caching” in a storage device refers to the use of high-speed volatile memory to collect write commands sent to data storage devices and cache them until the slower storage media (either physical disks or low-cost flash memory) can accommodate them. Most devices that use write caching require that power be supplied continuously. If you go to device manager it was found that you can also enable write cache buffer flushing (equivalent to the Intel AHCI driver) to get further performance increases. We’ll have to do some more research to find out more about this write-back caching that is happening and if it is a good thing or a bad thing. It seems that this would be a bad thing if your system weren’t on a battery backup, though.
The very first benchmark we ran showed an impressive 3238.9 MB/s, although we’ve never been able to duplicate that result again.