Over the past couple of years, at Broadband-Testing we’ve carried out a number of power consumption tests, primarily focusing on switches of the Ethernet variety.
While we found that there are considerable differences in the amount of power drawn by different switch products – for example, just a supervisor module in passive mode on a Cisco chassis switch can draw three times as much as a 24-port Gigabit switch by D-Link at full blast – in reality, compared with the amount of power used by the PC population of most companies, the savings that can be made at the switch level are not always significant.
So, we’ve switched (ahem) focus to power consumption of said PCs with a test of independent UK company, Certero’s PowerStudio 2.0 product, wot is being launched right now. The report will be released shortly on the www.broadband-testing.co.uk website (so watch this space, or keep visiting the website for an update) but meantime I can leak a few details of our findings.
First, it’s important to understand what Intel and others means when they describe each latest-generation PC product as being more power efficient than its predecessor. Yes they are, on a watt per CPU cycle basis, but the increase in performance each time as boggo PCs become multi-core, multi-CPU (in somes cases) beasties means that they actually draw MORE power overall than their forebears (a largely untold Goldilocks story and prequel to the famous one).
In a proverbial nutshell (think almond-sized), what PowerStudio does is to let you create a baseline measurement for total PC population power consumption (alliteration enters IT blogging!) and then create any number of power saving schemes to deploy and measure power saved (with lots of different metrics options) as a result.
Where the product gets really clever is that it can auto shutdown/wake PC clients remotely, using techniques such as Wake On Lan in combination with emails – very neat. Better still, it intelligently saves and reloads MS Office docs as well.
All this is great and means that the lazy user base can be controlled with lots of flexibility to suit any kind of business. However, the idea is for the software to act as an incentive for the users themselves to be more efficient and actually beat the software at shutting down their PCs. A “name and shame” list of worst users can be generated, as can a “star pupils” report, so we have a new employee competition to embrace. Obviously, this shouldn’t be taken too far with large bonuses offered, otherwise users simply won’t power their PCs and, instead, discuss TV for eight hours a day at the coffee machine and then get bonuses as a result of their power saving though, obviously, this has to be offset against additional power consumed by the coffee machine.
The reality is, however, that with a sensible set of power schemes in use and a well-behaved user base, cost savings can be absolutely huge. Moreover, with government schemes meaning that power consumption footprints need to be minimised to avoid punitive charges, there is every reason to look at something like PowerStudio and calculate just how much you can save by using it. Then, and only then, should you look again at power-efficient Ethernet switches. As that supermarket wot makes multi-billion pounds annual profits says, “every little helps….” It most certainly does in their case and it can in yours too.