The Tolly Group recently conducted an unusual test for Extreme Networks. The testing firm compared Extreme's switches to its competitors not on speeds and feeds - but instead on how much power the switches consume.

The Tolly Group is paid by the vendors for the tests it performs - in this case, Extreme. Extreme postulated that its BlackDiamond 8810 chassis-based switch consumed less power than comparable systems from Foundry Networks and Cisco, and looked to Tolly for verification.

In January, the Tolly Group tested the BlackDiamond against the Foundry BigIron RX-16 and the Cisco Catalyst 6509. It found that the Extreme switch used less than one-third the power, on average, of the Cisco switch, and half the power of the Foundry switch.

Tolly says it took power consumption measurements on the three devices while using more than a dozen traffic-load scenarios. The scenarios ranged from no load on the switch to a heavy load on eight 10 Gigabit Ethernet ports and 96 Gigabit Ethernet ports.

One example scenario: The Tolly Group sent bidirectional Layer 2 traffic at line rate across four ports of 10 Gigabit Ethernet on each box. The Extreme switch used 509 Watts, compared to 1,053 Watts on the Foundry switch and 1,811 Watts on the Cisco switch.

In the past, I've found that readers haven't been all that interested in "green" technologies. You're probably not even reading this now. But vendors keep harping on it, so someone must be paying attention - perhaps whoever is paying the electric bill. We recently put together an in-depth look at power issues in data center equipment - and its impact on the bottom line - for those who are interested. 

Tolly's executive summary of the test can be found here. 

Jeff Caruso is site editor at Network World.