Power And Efficiency
VIA’s claim to processor fame is probably the low advertised power draw, which has allowed several generations of product to operate without need of a cooling fan. On the other hand, efficiency is a measure of units of work per units of power, where weak performance can effectively cancel out any power savings. Is the Nano U1700 really a efficient as VIA claims?
Shuttle’s Intel dual-core platform drew only 33% more power than its single-core Nano solution. But besides having twice as many cores, the Atom is clocked higher. Will the Atom exceed the Nano’s performance level by more than 33%?
Indeed, the dual-core Atom performed 87% better on average than the Nano, even though only one out of five benchmark sets appeared to be dual-core optimized. We’ll probably have to wait for a 1.60 GHz dual-core Nano before we see VIA provide any real competition to Intel’s Atom 330.
The Atom slaughters the Nano in efficiency, in spite of VIA’s high-efficiency claims and low power consumption, and in spite of the ancient power-hungry 945GC chipset used in Shuttle’s Atom system. Even the microATX Celeron system, with its full expansion capability and 35W TDP processor, took a tiny lead over the Nano in efficiency.
maybe the performance delta would be the same but the usability would be more than "acceptable".
regarding the "benchmarking" stuff... i do not think that anybody sane would buy a nettop to use it for hardcore photo editiing or transcoding.
if i'd buy a nettop i'd be interested in several factors:
- to be powerfull enough for office work and maybe hd playback
- to use as little power as possible (which nano does nicely) as this kind of device would be rarely powered down
- to be silent (which nano is)
i think that performance per watt is irrelevant here as a system like this is not meant to be "performant". i think that the one that uses less watts, has more features is queter wins here.
my take on this is that via has a very nice platform and nano wins even if it's not the fastest.