if you used the cpu at full power for 6 to 8 hours per day for 3 to 4 years, you would pretty much eat up the initial savings in power costs.
I agree with all the power consumption concerns (not a fan of "more watts, less performance" myself), but I would like to point out that the same amount of money now (today) is
more valuable than an identical amount in 3-4 years (i.e., $100 now, vs $100 in 3-4 years)
If a business, for example, was considering two courses of action, A and B, where A incurs initial savings that will take 3-4 years to be 'nullified' due to overhead/maintenance costs (compared to B), and where B incurs initial extra initial costs, but will ROI in 3-4 years (compared to A), then NPV (and just general good cash flow management) will pretty much always side with choice A unless their lifespans were particularly protracted (15-20 years, such as real property, or maybe even a step down such as heavy machinery in production)
Therefore, if the metric in question is 'savings' (the maximization of which is the ideal), then any business/corporation with a good set of bean counters, if they were in this exact same quandary and had these exact same restrictions and central metric to consider to decide versus purchasing a Piledriver or an ivy Bridge, will end up choosing Piledriver. I do not wish to be misunderstood: again, this is only because we are within the strict confines of "which will save me money", and NPV and cash flow management (and no consideration of "minutes shaved off = directly more $$$$ in profit" or stuff of that sort because that was not in your exchange with Gikaseixas) will inevitably lead us to Piledriver
in this specific scenario. Even if the initial savings were to be "eaten up" in 3-4 years, that would still be the better course of action (again, consider the caveat above regarding protracted lifespans, so planning to hold on to the machines for 10 years+ will screw the computation)
In the real world, most use cases for Piledriver and Ivy Bridge will no doubt be heavily in favor of Ivy Bridge (personally, I cannot see how I would choose Piledriver over Ivy if i had to build a new rig from scratch, so bias alert there), but in this particular theoretical one, where the focus has become "savings" ($$) and the benchmark chosen is particularly close, Piledriver gets the nod.
Edit: Or better yet, I see now that I should have just used time value of money since we are only talking of ~$100 bucks, then I wouldn't have had to use the business scenario but still ended up at the same point.
Edit 2: This same reasoning is why I do not believe in the cost-effectiveness of buying "lower-power stuff" in the name of savings, when the effective wattage (as you have pointed out correctly earlier on, depends on actual operational hours) talked about is actually pretty small and will take years to realize. I am not talking of enthusiasts (we do crazy things coz that's what we do, and it's all for science and hobby, no rationale required, otherwise we'd all be using Celeron G550's), but again of the business case for "clearing out" last-gen equipment for "lower-power"/"energy-efficient" models being aggressively hawked by account executives of the respective vendors. The cash flow and cost accounting simply doesn't support it for any realistic / sane operational model and not the exaggerated ones they use when they try to scare us with their powerpoint sildes.
Facebook
Twitter