Doing the right thing
[HEXUS] Why not just leave it as Gold - surely you would sell more that way?
[Jim] We might sell more. Okay, we probably would sell more but that would be kind of, well, slimy, don't you think? Wait, I am a marketing guy... aren't we all slimy by definition [laughs]?
Jokes aside, we know we designed and built an 80Plus Silver PSU. We did such a good job that it is very nearly Gold, in terms of efficiency, but it wasn't designed to hit those targets so there is simply no way to guarantee that every PSU produced would meet that spec.
Look at it this way: when we market something, we're making a promise to the customer. He's giving us his hard-earned money in exchange for a product that he believes in, based on what we tell him about the product. That's a lot of trust for someone to put into a company and a product. I'd prefer to over-deliver on that trust because I think, in the long run, that's going to do more for Corsair than a few more sales of the HX850 or HX750.
You remember the Golden Rule, right? Do unto other as you would have them do unto you. It applies here. We're treating our customers honestly so that they will continue to trust us and keep buying our products.
[HEXUS] What certification did you feel your PSU would have got from internal testing?
[Jim] We know that the unit we sent passed at the 80Plus Silver levels in our test rig. We don't screen for specials units-if we did, we certainly wouldn't have asked to be reclassified as an 80Plus Silver unit!-but we do test the unit we send over to make sure it is functionally meeting our product specification.
[HEXUS] Is 80Plus ‘real world' testing or is it a simple validation - how does it differ to Corsair's internal testing.
[Jim] No, it's not "real world" testing, if you mean measuring efficiencies in a real computer, under actual conditions like playing a game. That's got even MORE variables in it than the testing we do for certification and it simply isn't practical.
Simply put, we both use a bench test with equipment that places a controlled 20%, 50%, and 100% load on the PSU and measures the differences between the input voltage and current, and the output, measuring how efficiently it is converted from 110/220V AC to 12V DC.
There are a lot of factors that vary in that process: temperature, manufacturing tolerances of the components, production variances (amount of solder, how much surface is in contact between components, variances in resistance within wires), etc. This introduces a lot of difficult to control variables, which means that the precision of the measurement between test rigs, between PSUs or even between days of the week (remember, temperature affects resistance!) isn't has high as people might believe.
That doesn't invalidate the results at all, but it does mean that it is fair to make sure you look at them carefully.