Now that we know what this router can do, let’s check it under real conditions and see if the efficiency of Atheros Super G technology is as high as promised.
I benchmarked the performance of the wireless interface in all four operation modes it supports: Super G with Static Turbo, Super G with Dynamic Turbo, Super G without Turbo, and full compatibility with the 802.11g standard (Super G disabled).
The wireless interface bandwidth was measured by the iperf utility with its default settings (TCP, 8KB window size). Each test was run for 2 minutes. An AVI file was used as the source of test data (that is, the LZW compression employed in Super G mode didn’t affect the results).
To test the TEW611BRP router I took a notebook with two network interfaces (wired and wireless) between which the data was being transferred via the router. The notebook’s wired interface (SiS 900, 100Mbps, full duplex) was connected to one of the router’s LAN ports. The wireless interface was provided by a D-Link DWL-G650M adapter based on the same chipset as the TEW-611BRP and thus supporting all the exclusive technologies from Atheros. I used the latest version of the adapter driver available at the time of the test.
Here are the test results:
I guess the diagram doesn’t need my comments. Super G technology really works and it works very effectively. Even using Super G without the channels combining feature results in a 40% increase in data-transfer speed over the wireless interface (in comparison with the standard 802.11g mode). When the doubling of channels is turned on, the performance is about 2.8 times the standard speed.
Yes, this impressive result was achieved under ideal conditions: there were no devices in the test network other than the router and the wireless adapter that would affect the data-transfer speed (this is why the results of the Static and Dynamic modes almost coincide). Besides, the router and the notebook’s wireless adapter were very near each other and this helped reach a maximum performance due to lack of radio interference. In a real network, the overhead associated with providing compatibility with Super G incompatible devices, with error correction and even with sharing one access point among several adapters may prove quite big. And still, this test shows that you can achieve a considerable increase in the performance of your office wireless network by using a TEW-611BRP router with Super G compatible adapters (as for a typical home WLAN – it doesn’t differ much from our test network).