5200+ X2 Windsor @ 3.2Ghz
Biostar TForce 570 SLI
2x1Gb OCZ Titanium PC2-6400
3x74Gb WD Raptors RAID0
I chose several benchmarks to compare average read/write times, including HD Tach 18.104.22.168, HD Tune 2.54 and the SiSoft Sandra IIc Physical Disk Read test. I also used a single 600MB file and a collection of ~500 files totaling 800MB for a "real world" read/write series of tests.
First up is the USB 2.0 interface.
All of the benchmark utilities top out in the mid- to upper-30MB/s range. You can tell especially by the HD Tune results that the USB interface is the limitation here. The drive is able to keep the bandwidth saturated and the throughput over time maintains a steady 35MB/s. This is a typical result from other USB drives I have reviewed in the past. In the real world tests you can see that the write speed also tops out around 30-35MB/s with the single large file taking advantage over the multiple smaller files. Read speeds are much faster which brings the read/write average up considerably.
Now let's run the eSATA interface through the same battery of tests.
Wow, what a difference the eSATA connection makes. Although throughput averages out in the mid-40MB/s, it is immediately apparent from the HD Tune result that now the drive itself is the bottleneck. Speed starts out around 55MB/s and tapers off just as any single internal drive would. Being that this is an older SATA-150 drive we really can't expect to get much more sustained bandwidth from it. Certainly a faster drive would yield better results, with eSATA there's no reason why you can't get the same performance outside the case as inside.
On to final conclusion.