Testbed and Methods
Our testbed was configured as follows:
- Processor: Intel Core i3-2100 (3.10 GHz, 2x256 KB + 3 MB)
- Mainboard: Socket 1155 Gigabyte GA-Z68X-UD3P-B3 (Intel Z68)
- Hard disk drive: 3-terabyte Western Digital Caviar Green WD30EZRS 5900 RPM (SATA)
- System memory: 1 GB DDR3 SDRAM SEC M378B2873GB0-CH9 (PC10600, 1333 MHz, CL9)
- Graphics card: HIS HD 3870 IceQ3 Turbo H387Q512NP (Radeon HD 3870)
- Power supply: Seasonic M12D SS-850EM (850 W)
- Operating system: Microsoft Windows 7 Ultimate 64-bit SP1
We installed the latest drivers from the manufacturers of our components. The tested disks were NTFS-formatted with the default cluster size.
The tested USB drives were connected to the mainboard’s USB 2.0 (Intel Z68) and USB 3.0 (EtronTech EJ168A) ports.
We used the following benchmarking tools:
- FC-Test 1.0 b13
- CrystalDiskMark 3.0.1
There were three FC-Test patterns corresponding to three typical usage scenarios:
- Writing and reading large files (we use a single 900MB file) such as videos, software distributions, etc.
- Writing and reading medium-sized files (2 to 12 megabytes with a total size of about 1 gigabyte) which is comparable to copying photos or MP3 music.
- Writing and reading files of an installed program (a folder with Microsoft Office 2007). Although this is not a very popular scenario, it is the most difficult one for flash drives as it includes a lot of small files.
We also used CrystalDiskMark for you to be able to compare your results with ours if you want. We ran each test of that benchmark five times using 1000 megabytes of test data.