Associate
- Joined
- 31 Oct 2012
- Posts
- 2,243
- Location
- Edinburgh
Well what people might be referring to is that if it is only 390/970-390x/980 performance, the 970 had a 145w tdp and the 980 had a 165w tdp, and that was 20 months ago.
If the tdp of this is only under 150w even the perf per watt isn't all that wonderful.
Obviously if AMD do manage to get it under 100w for that level of performance, then yes it will be good.
TDP can't be compared across manufacturers as it's not a standardised measure and indeed NVIDIA use the term to mean something different to the rest of the industry. Intel & AMD both define it as the amount of waste heat that a cooler must dissipate in order to keep a silicon chip within its operating temperatures. NVIDIA view TDP as an input. There are no standards at all for what to run to determine it etc which has led to some very misleading results over time (including some gens where NVIDIA were closer to the actual power used and some where AMD were).
You can only look at power consumption tests to compare cards - which is what I'm sure we'll do once the 480 is released.
Gotta be careful though with this as most review sites don't actually measure this, they measure whole system power at the wall then either present that (giving massively inflated numbers) or guesstimate the power draw of the rest of the system based on something they think is 'known' resulting in pretty variable / made up numbers. Sometimes they even remember that the PSU isn't 100% efficient but sometimes not.
Some sites do use proper risers & modified aux power & look at power draw over a period with short sample intervals etc but even then you've got a few factors such as the risers causing different airflow resulting in different temperatures hence different efficiency, as well as game or program choice & settings used causing the results to vary significantly (along with CPU & RAM choices of course - do you try to 'minimise bottlenecks' or do you try to get a picture of it's likely power draw in a system a user may actually have, as these may be substantially different).
I've kinda forgotten what I set out to write - basically that reviewing power draw well is difficult and subjective I guess? Sorry, mind wandering...