A lot has happened in the USB 3.0 world since we took a look at Kingston’s first ‘Ultimate 3.0′ drive this past fall, and thanks to some progressions, the company has just released a ‘G2′ variant of the drive. It’s identical in looks, it’s cheaper, and it proves to be the fastest non-SSD USB 3.0 device we’ve tested to date.
Over the past year, many storage companies have been jumping on the CrystalDiskMark bandwagon to help strut their product’s performance, and it’s easy to understand why. Compared to others, CrystalDiskMark delivers results that are much higher, and in some regards, they could be considered unrealistic given that real-world tests and even other synthetic benchmarks never seem to back up its claims. We include it for the sake of interest and because it is still a very thorough benchmark.
As expected, the G2 drive easily outpaced the G1 in this test.
While CrystalDiskMark does well to show the absolute top-end value of a storage device, ATTO doesn’t fall too far behind, if at all. Its test uses a wide-range of cluster sizes, for both read and write, but we only note 4KB, 64KB and 1024KB of the former. For those interested, we use a queue depth value of 10 for testing.
The theme continues here, with minor gains seen with 4K reads but significant gains seen with the others – at least, in USB 3.0 mode.
For real-world testing, we use a set of files and folders for the sake of measuring transfer speeds, and also convert images and music on the storage device to see just how well it fares for large intensive operations. For the transfer speeds, we use both 4GB and 16GB files and folders, and for the former, we also perform copy tests, which refers to copying the file or folder on the storage device. We don’t do this for our 16GB files and folders as some 32GB drives refuse it due to coming so close to the total density.
Looking at the USB 3.0 results, the performance gains seen with the G2 drive are enormous, cutting a 4GB file transfer down from 63.51s to 52.74s, and even bigger gains can be seen with the file copy. However, the lack of total IOPS can be seen as a detriment with the folder tests, where thousands of files are being transferred. The G2 preformed noticeably worse in this regard, taking 237.63s to transfer a 4GB folder on the G2 to only taking 109.43s with the G1.
The performance loss on the G2 isn’t quite so severe with our folder test here as it was with the 4GB test, and where single files are concerned, the G2 is far superior.
With our file transfer tests leaving a bit to be desired on the folder transfers, the results seen with Adobe Lightroom and dBpoweramp are refreshing. This sure seems to be a drive that’s a master of one, rather than a jack-of-all-trades.