Setup and Testing
Test platforms: Giada i35G ATOM D2500 based MINI PC, AMD E450 HTPC and a Gigabyte Z77X-UD5H with an i5 3570 CPU.
Test Drives used in the Blizzard: Seagate 7200.10 250GB SATAII HDD and Mushkin Chronos 120GB SATAIII SSD
Comparison external drive bay: Thermaltake BlacX 5G (Using Seagate 7200.10 250GB SATAII HDD and Mushkin Chronos 120GB SATAIII)
Comparison external drive: WD MyBook 2TB usb2.0 drive
USB 2.0 tested on the D2500
Using ATTO 32 Disk Bench we find no measurable difference on the USB 2.0 interface.
We then move to SSD testing on the USB 2.0 interface with no real changes in our results
What we learn from these tests is that the USB 2.0 Bus is our limiting factor. We did these tests for two reasons – first is to check for deficiencies in the control mechanism of the Icy Dock Blizzard and second to show the limitations of the USB 2.0 bus.
USB 3.0 tested on the Gigabyte Z77 Based board
On the left are the results from the Icy Dock Blizzard riding the USB 3.0 Bus and on the right are Thermaltake BlacX 5G results using the exact same SATA II 7200 RPM hard drive. 
Next we have USB 3.0 testing with a SATA III SSD and, as above, the same drive was used in both external drive adapters – Icy Dock on the left and Thermaltake on the right
eSATA III tested on AMD E450 Based system
testing we had some anomalous write results that we just can not fully explain using a Marvell eSATA controller and a Solid State Drive. We decided not to let this ride as it is not a fair representation of the product. When we were finished with the testing – we set up another testbed, reran the eSATA tests using a different brand controller and everything went smoothly. Results were par in reads and writes with internal drives. 
We must conclude that our initial assumption about the weak write performance on SSDs was due to the Marvell controller. We include this information because it may not be an isolated case and we feel that it is our job to report findings that may be relevant to our readers.
