Performance Testing
The following two systems where used for the next two benchmarks.
| USB 3.0 | |
| CPU: | Intel Core i7 2700k |
| Motherboard: | ASUS Maximus IV Extreme |
| Ram: | G.Skill RipjawX 2133 11-11-11-31-1t |
| GPU | ASUS DirectCU 460 GTX |
| HDD(boot) | OCZ Vertex 2 50gb |
| HDD(storage) | Western Digitial RE4 1tb |
| Wi-Fi | |
| Laptop: | ASUS G73JH |
| CPU: | Intel Core i7 720QM |
| Ram: | Kingston ddr3 1600 9-9-9-24-2t |
| GPU: | ATI Radeon 5870m |
| HDD(boot): | Seagate MomentusXT 500GB 7200rpm |
| HDD(storage): | Seagate MomentusXT 500GB 7200rpm |
USB 3.0 testing
Testing for this was pretty straight forward. The drive was quick formatted and then bench marked. Rinse and repeat.
ATTO, Stock Settings
ATTO, Queue Depth 10
CystalDiskMark, 64bit
When you look at all three of these benchmarks they are surprisingly close to each other for the overall results. Not far off what one would expect out of a spindled mechanical hard drive over a USB 3.0 interface. Part of this has to do with data handling before it makes its way through the USB interface to the drive.
Without Having another external USB drive to compare it to it does seem kinda out of perspective. We’ll clear this up in a minute with a nice big comparison of the performance of different interfaces, across two separate drives.
Wireless Testing
As we jump into the wireless testing portion of the benchmarks things get a little more complex. So I am gonna do my best to guide you through the test process as well as the data formatting. Since we don’t really have a baseline out there for comparison, some additional data points have been put into the table as reference points.
Since the majority of our viewers are use to seeing CrystalDiskMark from our SSD reviews. We’ve used this to test both wireless and wired performance of the drives. This has been done by mapping the the networked device to the machine. So that you can access it from CrystalDiskMark. This gives you a sort of generalized idea of the performance your getting.
Network and wireless testing in general is somewhat of a dirty thing. There more of a variance in the performance. Wireless networking in particular has a extremely high loss/error rate. This leaves the actual performance of most devices to be much lower than the theoretical maximum of the medium. Increased distances, EMI and other other physical obstacles can add to this. Wired Networking is also effected by this to some degree though it generally easier to have consistent performance across the wired medium than a wireless one.
Software and operating system also plays a huge part in network performance. Depending on the versions of operating system in question, and settings used for file serving. In a ideally optimized environment the software will fill in for hardware short comings such as a slower hard disk. This is not the case for what we are gonna called standard real world application.
Our tests are being done across either one or two fully patched windows 7 ultimate boxes. With the Laptop being the workstation in question. For the wired networking tests the drive is connected to the listed core i7 desktop, and shared through a gigabit switched network. For wireless testing the drive is placed three feet from the laptop. This is a little further than would typically be the case for a single user setup. However since this isn’t target at a single user in the wireless environment this is a nice intermediate distance for what you would find in a meeting or work place.
Being that the Gauntlet Node is setup to provide streaming to five users at once. Its network performance overall is setup to a maximum of roughly 2/3rds the theoretical maximum of the device (150mbps). This value is then split into fifths so that it can support five users. This results in approximately 20mbps per connected and streaming device. Which is what we observed as its maximum sustained peak. At Least this is the math we are using to explain whats going on.
So after looking at the data all laid out nice you’ll notice a few trends. The Drive we chose in question is and can be a performance bottle neck for standard computing. However when used in the USB 3.0 Enclosure the software really tends to make of for the normal shortcomings of the otherwise performance limited drive. It also has the benefit of being more than suitable for wireless usage. As the drive we used is not gonna be a bottle neck for single user situations. It might be for a five user setup. Mostly likely not, and the simple solution should this ever become a problem for you. Put a faster drive in.
