cmon jmac, we all know that stuff!! Do you think we've not read the reviews as well
With our magnificent tech reporter NegSun, such tidbits will never escape us!
Originally Posted by teehee@Toms
However, we used the SYSmark 2007 Preview suite as it represents the majority of workloads applicable to a wide user base, and also because it does include some idle time, which we believe reflects real user behavior pretty well.
"we" ?? The people at Toms are qualified to conclusively report on long term effects of consumer energy consumption profiles?? I haven't come across any report of a longitudinal study design
of theirs to test this.. I doubt we ever will.. people who soil themslves at TRI-SLI setups and glowing case fans are less likely to be concerned about this issue at any further depth than a SYSmark test.
There are a number of initiatives research centres/energy providers are undertaking to look into this type of stuff. One of the most notable being Project NOAH(University Of California, Irvine)
of course. There's some great articles analysing their methodology and the unique "AppTrack"
software they developed.
One of the challenging aspects(from a research POV) of the IT world is the high rate of change for new hardware, due to ever improving engineering standards. This is perfectly summed up by Moore's Law
which any hardware geek worth their salt will know intimately
A contiguous research model is something that is necessary to keep up with all this, which is an expensive undertaking for governments and energy providers, who are skilled at penny pinching
Now, I wonder if "Tom's Hardware" can meet that challenge instead?[/bollocks they can]
What it boils down to:
True, representative data about energy consumption for home computer use is complex to gather, requiring a specialised research methodology that will not only gather the data correctly, but also keep up with ever evolving computer hardware standards. The information reported in small scale hardware reviews, whilst interesting to read, can in no way approximate an adequate research design, neither in scale nor intention.