QUOTE(eDwanD @ Apr 29 2017, 12:21 AM)
» Click to show Spoiler - click again to hide... «
This is from tweaktown (3D marks and Metro Light). Yours from tomshardware (gaming and tortune).
Tourtune test in tom's hardware result more power consumption than rx 480 nitro.
If tweaktown is true, 27w~37w increase just for a gpu is not a performace/watt choice. Anyway, the performace/watt oledi lose to gtx1060 long time ago @@
Who's right? I dunno
If there's no increase of power consumption, its great, actually, even there is increase of 30w average, wont hurt u tnb bill so much per year @@
If really willing for save tnb bill, just go to gtx 1060 @@
But still i wont happy, RM200 from nitro 480 to nitro 580, price increase 20% to get u 5% overall performance increase, 10% for game optimization.
I prefer direct measurement instead of "total system".
Also i don't understand, they tested it with Ryzen R5-1600x?
QUOTE
For our mid-range testing, we're shifting over to testing the cards on our AMD Ryzen 5 1600X system. It represents more real-world results compared to benchmarking them on the Core i7-7700K, especially when the Ryzen 5 1600X is priced at $100 cheaper than the 7700K, and offers 6C/12T of performance over the 4C/8T chip from Intel.
CPU: AMD Ryzen 5 1600X (6C/12T)
Cooler: AMD Wraith Cooler
MB: ASRock AB350 Gaming K4
RAM: 16GB (2x8GB) GEiL EVO X 3200MHz DDR4
SSD: 1TB OCZ RD400 NVMe M.2
PSU: be quiet! Dark Power 1200W
Chassis: In Win X-Frame
But at OC and Power page they said
QUOTE
With the overclocked Polaris 20 core hitting 1505MHz in my testing, the SAPPHIRE Radeon RX 580 Nitro+ sitting on the Core i7-7700K system was using an average of 270W, while peaking at 290W.
And their Power Consumption graph show the card averaging(?) ~280watts at stock? with stock Ryzen R5-1600x system?
» Click to show Spoiler - click again to hide... «
But their Ryzen R5 review show R5-1600x system use similar power to i7-7700k system?
» Click to show Spoiler - click again to hide... «
» Click to show Spoiler - click again to hide... «
They mentioned midrange card and adjust their test system for "midrange build" but wiling to waste their time to test on 4k, 5k and even 8k resolution when giving more insight about the card frame time performance at 1080p/1440p is more valuable info to viewers/readers.
IIRC, Toms torture test = Furmark and gaming test = Metro:LL@4K. Like i mentioned few pages back..
QUOTE
AMD card power consumption varies depend on BIOS and chip binning
that's what i learnt when i'm still using 290x, i think it still apply to Polaris line up.