The green debate in IT may have taken something of a back seat in the face of wider economic concerns, but it doesn’t stop certain gems appearing that highlight the problems in the area.
Recent research covered widely in the popular press said that a Google search could take as much power as boiling half a kettle of water, producing as much as 7g CO2 in the process.
Were that to be true, it would mean that every time we typed something into Google to do a search, we were essentially making a lot of CO2 that would need to be offset in some way to avoid the all too familiar results. A poor lookout for those enthralled by “Google Whacking”!
However, Google were quick to respond and said that the research was flawed and that the real number was closer to 0.2g of CO2 produced and a fraction of the power consumed.
Now, I do not have the wherewithal to determine which system of assessment is correct, but the problem highlights what we, the journalists behind Mediateam’s technology titles, have been on about for quite some time. This is the fact that without some standardised method of assessment for determining the power usage of IT equipment, despite its myriad configurations and usages, that would give an accurate figure for the way that power is consumed and thus the impact it has for the environment, there is no way of settling such arguments.
So, if Google used equipment that was energy labelled, and if there were standardised methods for determining how that equipment uses power as it operates under a given set of conditions, then there would be no debate, no claim and counter-claim to save face. There would simply be a set of figures that would be accurate enough to allow a basis for comparison with other operators in this area, and thus some measure of wisdom as to whether the operation in question, internet searches, were being done as efficiently as possible.
However, the reality is that there is no standardised energy labelling of servers and network switchgear, nor even an accepted method of evaluating a dynamic load. As a result, these kinds of debates can emerge with potentially damaging consequences as in this case, Google felt compelled to defend itself against what can be reasonably termed negative press.
But, Google is not entirely innocent here in that it does build its own systems for servers rather than buying readymade options. The impact of this is that without sharing the specs, outsiders can’t really determine how the things work. Give away trade secrets in the name of honesty? Perhaps not, but for as long as this situation remains, where there is no definition for energy accurate consumption for IT equipment in dynamic circumstances, then these storms of debate will continue to be of greater incidence and effect. A bit like the weather really.
Subscribers 0
Fans 0
Followers 0
Followers