Testbed and Methods
Our testbed was configured as follows:
- Processor: Intel Core i5-2500K (3.30 GHz, 4x256 KB + 6 MB)
- Mainboard: ASUS P8Z68-V (Intel Z68)
- Hard disk drive: 1-terabyte Western Digital Caviar Green WD10EADS 5900 RPM (SATA)
- System memory: 2x4 GB Kingston HyperX KHX1600C9D3X2K2/8GX (PC12800, 1600 MHz, CL9)
- Graphics card: Sapphire Radeon HD 6970
- Power supply: XFX Black Edition PI-850B-NLG9 (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 not reformatted; we used the file system and cluster size chosen by their makers.
Each HDD was connected to the mainboard’s back-panel USB 3.0 ports based on an ASMedia ASM1042 controller.
We used the following benchmarking tools:
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 external disks as they have to deal with 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.