Performance Testing Methodology

Testing an external storage array is not as simple as putting together a bunch of files, dragging them onto the arrays drive folder in Windows and using a stopwatch to time how long the transfer takes. Rather, there are factors such as read / write speed and data burst speed to take into account.

For these tests I used a combination of the ATTO Disk Benchmark, HDTach and the SIS Sandra Removable Storage benchmark. For all USB & 1394a testing an ASUS M2N32-SLI Deluxe motherboard was used, with it’s built in USB and FireWire controllers. For all 1394b testing a PCI daughter card with a Texas Instruments controller chip was used.
All tests were run 4 times and only best results are represented.

For information purposes here is the theoretical maximum each connection is cable of:

USB 2.0 = 50MB/S (480Mbit/s)
FireWire 800/1394a = 40MB/s (400Mbits/s)
FireWire 800.1394b = 100MB/s (800Mbits/s)

Theoretical numbers aside, FireWire is a much less resource intensive connection based on “Peer-to-Peer” architecture where the device itself is “intelligent” and handles most of the data transfer and data collision work. USB relies on the older “Master/Slave” architecture where the host computer has to do all the necessary work.

This means that more of USB’s throughput is used just to keep the connection alive (AKA overhead), this leaves less useable bandwidth for information. This is why a theoretically slower connection (1394a) will usually have better speeds than its USB counterpart.

Complete Test System:

Processor: AMD X2 4800+
Motherboard: ASUS M2N32-SLI Deluxe
Memory: 2GB Corsair TwinX PC2-6400
Graphics card: Evga 7900gtx 512mb
Hard Drives:
2x Seagate 7200.10 320GB (used for enclosure testing)
1x Western Digtal Se16 500GB
1x Western Digital Raptor 150GB
Power Supply: Enermax Liberty 620W
Case: CM Stacker 830 w/ 8 Scythe E fans
Firewire PCI Card: PPA Int'l 1225 FW800 card