Quote from: sgbett on May 18, 2013, 04:02:36 AM

Quote from: Branksy on May 17, 2013, 07:40:53 PM

My sources are irrelevant. What is relevant is that nobody has addressed my points with any valid counter arguments, or god forbid a link and not just a wall of text with no factual statements.

Your point(s) all seemed to be that you and your friends had some opinion on what should or should not constitute "performance", and that your opinion was that power consumption didn't count.

Well here is my opinion:

Performance describes the manner in which something functions. Those attributes that are particular to its operation, how it 'performs'.

Specifications describe what something is. Its physical attributes, its form.

Now you said...

Quote from: Branksy on May 06, 2013, 02:58:20 PM

As i'm sure others here do, i work professionally with a lot of electrical engineers. I specifically asked a few before betting if power requirements would be considered performance or just specifications. What i heard back unanimously (from my small sample) is that performance doesn't relate to power consumption, that would be Performance Per Watt, which is different. It's the same as saying that case size relates to performance. Energy is just a utility to achieve performance, not the performance itself.

I understand that the reason BFL looked so good is because of the performance per watt, but just because we want low wattage for maximum profits, and just because that's what we based our pre-orders on, still doesn't mean it's relevant to "Advertised Performance".

Lets break that down:

You contrast performance and specification as being two distinct things. This statement is important because it sets the mood. You will use this idea later to affirm the disjunct.

You ask your friends to categorise "power requirements" as performance or specification. However "requirements" are something that might form part of an objects design, so naturally it would seem fit that requirements are considered part of a device's specification. Requirements exist independently of an object ever operating.

When phrasing their response, you play a clever trick and switch the term "power requirements" for "power consumption", you say your friends tell you "power consumption" is not an aspect of performance.

Power consumption is something that happens when a device is functioning. Consumption happens (more than likely in line with the specified power requirements) when the device 'performing'. It is an aspect of 'performance', which will likely be in line with the specified "power requirements".

You try to redefine this aspect of performance as being something else. However in creating this other "not-performance" category, you actually use the word performance to describe it!

You then make the statement "It's the same as saying that case size relates to performance".

This is where you attempt to close the deal.

You attempt to assert that "power consumption" is not performance, and use your original premise of categorising things as "performance or specification" to assert that it must then be part of the device's specification.

"performance per watt" IS performance. Saying "GH/s isn't performance, it's performance per watt" is nonsensical.

It is like saying "Mallards aren't birds, they are ducks" or "Ferrari's aren't vehicles, they are cars".

Performance per watt is not performance, it's efficiency. I pointed that out several pages ago. The only people who don't seem to understand this are scummy little weasels who work in finance. Ask anyone who works with technology, ask on a forum such as Anandtech's CPU or GPU forum and they will tell you the exact same thing. Performance is about speed, Performance per Watt is about efficiency.

It should be blatantly obvious to everyone who isn't brain dead that performance per watt is not the same as performance, otherwise we would simply call it performance and not performance per watt.

Of course you're going to defend their moronic, biased decision, you benefited from it.

How about we get some advice from people who live and breathe technology instead of relying on scummy little rats who live and breathe finance?

Performance per Watt is efficiency, not performance. It's that simple.