PCPer AMD R9 Nano review

[Mahigan]

The first revision of Fury-X had a bad pump. True. The second revision fixed that issue.

Personally, I dislike AIO watercooling.

[zMeul]

Oh dear (this is turning into a joke now), zMeul, you just linked a video you didn't watch? (Your thread opening, PCPer video)

Or you simply lack the ability to fucking UNDERSTAND. That's what you're showing off here constantly. You understand nothing and you're just deaf and blind at the same fucking time.

Look. The R9 Nano has 2 limits.

I watched, but that doesn't mean I have to agree with Ryan's explanation or interpretation of the data

 

here: http://www.tomshardware.com/reviews/amd-radeon-r9-nano,4285-9.html

 

game:

FurMark:

 

notice anything? during the Furmark stress test, both the peak and the average power draw (video card only) is lower than in the game stress test

what's the one common factor? temperature:

 

conclusion: the R9 Nano hits the thermal limit before it reaches the power limit

as for the 85deg limit, that is set when the power limit is set to maximum (+50%)

---

funny how this 175W limited card draws over 400W max and has 180+W average

[QueenDemetria]

notice anything? during the Furmark stress test

Furmark overheated yet another graphics card? Tell me more...

 

Seriously, Furmark is a completely unrealistic test and should only be used for testing "theoretical limits", not actual limits of cards. And if you want to know why Furmark made it hit a thermal limit rather than a power limit, do your research on "Furmark Power virus", over the years ATI/AMD and nvidia got tired of this stupid program wrecking graphics cards so some cards reduce the power once Furmark is being ran(again, another reason why Furmark should not be used for comparison sake)

 

 

ATI: http://www.geeks3d.com/20090916/furmark-slowdown-by-catalyst-graphics-drivers-is-intentional/

 

And nvidia: http://www.geeks3d.com/20101109/nvidia-geforce-gtx-580-the-anti-furmark-dx11-card/

[Dabombinable]

Furmark overheated yet another graphics card? Tell me more...

 

Seriously, Furmark is a completely unrealistic test and should only be used for testing "theoretical limits", not actual limits of cards. And if you want to know why Furmark made it hit a thermal limit rather than a power limit, do your research on "Furmark Power virus", over the years ATI/AMD and nvidia got tired of this stupid program wrecking graphics cards so some cards reduce the power once Furmark is being ran(again, another reason why Furmark should not be used for comparison sake)

 

 

ATI: http://www.geeks3d.com/20090916/furmark-slowdown-by-catalyst-graphics-drivers-is-intentional/

 

And nvidia: http://www.geeks3d.com/20101109/nvidia-geforce-gtx-580-the-anti-furmark-dx11-card/

Furmark weeds out the cards with shit quality components from the good ones. Simple.

[QueenDemetria]

Furmark weeds out the cards with shit quality components from the good ones. Simple.

Yep, thats why I use it(if it cant last 24 hours with furmark it aint good enough for me), but for people wanting to compare temps and power draw with furmark any argument is instantly invalid as soon as they say "I used furmark". AMD/Nvidia have limits set on these programs, so its no wonder why these results are wonky.

 

Edit: Furmark shouldn't be used for benchmarking is what I'm saying, although stress/endurance testing is fine as long as you don't mind risking the hardware.

[Mahigan]

Furmark????

Get out of here! You think a gamer is going to be playing furmark?

Now you're just grasping at straws.

[kacper]

Is there any logical explanation why Fury/X/Nano don't have HDMI 2 ?

 

I can understand R3brand series but new products?

[LucidMew]

Is there any logical explanation why Fury/X/Nano don't have HDMI 2 ?

 

I can understand R3brand series but new products?

 

It's because the Fury* series of cards are essentially a beefed up tonga 285 with HBM, and the [thing that drives the display output] just can't do HDMI 2. 

 

It's a limitation of the card, not AMD being stupid and purposely gimping their product.

[marldorthegreat]

so it thermal throttles - the rumors were true  :lol:

Nope. it does not. It power throttles when it reaches 175 watts. it never gets abouve 80 celcius

[marldorthegreat]

It's because the Fury* series of cards are essentially a beefed up tonga 285 with HBM, and the [thing that drives the display output] just can't do HDMI 2. 

 

It's a limitation of the card, not AMD being stupid and purposely gimping their product.

I think it is also to do with that hdmi 2 is a bit arkward and is expensive

[marldorthegreat]

funny how this 175W limited card draws over 400W max and has 180+W average

every card does that BECAUSE IT TURBOS

[zMeul]

every card does that BECAUSE IT TURBOS

then, you don't have a 175W power limit - boost or not

you do not get to have 180+W average power draw when you have a 175W power limit

boost?! what boost?! the clocks barely reach 900Mhz, not even the rated boost clock of 1000Mhz (!!!!) - the clocks average at about 875Mhz

I'd seriously like to see this card water-cooled to put this matter to rest

if what I say is utter shit, the water cooling would make absolutely no difference to clocks in the exact same game

[Sakkura]

funny how this 175W limited card draws over 400W max and has 180+W average

 

Only in short spikes. The power limit is apparently applied to sustained power draw.

 

175W isn't the limit, it's the TDP. The limit keeps the card pretty close to the TDP; the power leakage increases when the card heats up, so when it's cold it's about on target for 175W, but once it heats up it's 180W plus change.

[Popingheads]

then, you don't have a 175W power limit - boost or not

you do not get to have 180+W average power draw when you have a 175W power limit

 

 

Thats funny I don't remember you making an issue of this when the 980 was released with a 165 watt power draw yet commonly pulled 185 average with spikes up to 300 watts.

 

 

Also the Nano is a little over 175 because of efficiency reasons I remember reading, when the card is cold the power draw doesn't exceed 175, only after it heats up does it get a little worse. The limit is still 175 watts but that doesn't include power losses in other components like the VRMs, which get slightly worse at higher temperatures.

 

 

notice anything? during the Furmark stress test, both the peak and the average power draw (video card only) is lower than in the game stress test

what's the one common factor? temperature:

conclusion: the R9 Nano hits the thermal limit before it reaches the power limit

 

 

I'm lost on how you reached this conclusion too since the temperature graphs clearly show the Nano stays cooler during the Furmark test, meaning temperature is not the reason the card draws less power, rather the temperature being lower is a product of the card using less power in the Furmark test. I don't see how this data at all correlates to the hypothesis you are putting forward that the card is thermal limited.

 

Also the power limit setting in CCC does not affect the thermal limit of a card either.

[ivan134]

Furmark????

Get out of here! You think a gamer is going to be playing furmark?

Now you're just grasping at straws.

You just don't like Furmark because you're a COD player and don't like games with good character development.

[Mahigan]

You just don't like Furmark because you're a COD player and don't like games with good character development.

 

I'm a (whatever Blizzard releases) player and a huge fan of the Battlefield series

 

On a side note.. I miss Age of Empires.

[QueenDemetria]

Is there any logical explanation why Fury/X/Nano don't have HDMI 2 ?

Honestly because it doesn't need it. So HDMI 2.0 is mainly relevant for 4K, which if you are using a monitor you use DisplayPort or if you are using a TV you use HDMI. The problem is that the amount of people that own 4K TV's is relatively small(a majority of people are still on 1080p) and  the amount of people that use their 4K TV for PC gaming is even smaller. By the time that couch PC gaming becomes mainstream, the R9 300 series and Fury's wont even be relevant anymore, so its not really worth AMD's efforts to put a technology into a card that only a few people would actually use.

 

then, you don't have a 175W power limit - boost or not

you do not get to have 180+W average power draw when you have a 175W power limit

Ahh yes, another person that I have to explain efficiency to. So electronics don't "just use" X amount of watts(unless if you calculating the whole system which then it does, but  we are about to do this). So when AMD says its a 175w card, it very well might be, but it uses slightly more than that(because lets face it the chart you provided showed a max 11w difference, which is hardly anything) it means that whoever was testing didn't account for the power loss with the power supply. You know the 80 Plus rating, we've all heard an used it. Well on a basic level in order to get that rating your powersupply needs to be able to effectively convert at least 80% of the power from the wall and make it usable for your components. Well, powersupplys have gotten good, so good they are reaching 90-95% efficiency, meaning that there is only a 5-10% power loss with all of the power conversion. Well now lets do some math:

Worst case scenario(90% effficiency, 10% loss): 175 x .1  = 17.5 <-- Of watts lost during the power conversion

 

Ideal scenario(given that the tester is using a great PSU, with 95% efficiency and 5% loss):

175 x .05 = 8.75 <-- Of watts lost during 

 

So what that means is that by the time the 120v from the wall(since Toms is US) is converted to the 12v for the GPU, it will either use an additional 17.5w to 8.75w(with current technology), so the R9 Nano will use anything from 175w to 192.5w(assuming 100 to 90% power supply efficiency).

 

If you think I'm making this crap up, re read the toms article, they even note the 8w difference and state: "we actually hit this number(when referring to the previously mentioned 175w)". So there, not only did I debunk the power limit(twice now?) but so did "Tom".

[Prysin]

Digitalfoundry claimed their Nano was silent.

these reviewers often pass their cards around... so some reviewers often review the same card...

 

say 20 cards go out for review and 5 has coilwine... but those 5 is passed around to atleast 3 sources... thats 15 bad mentions just cuz 5 cards went into circulation...

[Nena Trinity]

these reviewers often pass their cards around... so some reviewers often review the same card...

 

say 20 cards go out for review and 5 has coilwine... but those 5 is passed around to atleast 3 sources... thats 15 bad mentions just cuz 5 cards went into circulation...

More or less that sounds about right! :3

[Sakkura]

Honestly because it doesn't need it. So HDMI 2.0 is mainly relevant for 4K, which if you are using a monitor you use DisplayPort or if you are using a TV you use HDMI. The problem is that the amount of people that own 4K TV's is relatively small(a majority of people are still on 1080p) and  the amount of people that use their 4K TV for PC gaming is even smaller. By the time that couch PC gaming becomes mainstream, the R9 300 series and Fury's wont even be relevant anymore, so its not really worth AMD's efforts to put a technology into a card that only a few people would actually use.

 

Then again, the R9 Nano is a niche product for the few. It seems to be particularly aimed at people who want high-performance HTPCs, and an obvious example would be someone who wants to game on their 4K TV.

 

That said, redesigning the display controller for just this one niche product would not be worth it for AMD. It's an unfortunate limitation for the R9 Nano, but for most of their other cards it doesn't matter much.

[ChrisxIxCross]

ITX systems are such a small niche part of the market their is absolutely zero reason for the Nano to exist. Not to mention most full size cards can fit in micro ATX cases anyway.

[HalGameGuru]

ITX systems are such a small niche part of the market their is absolutely zero reason for the Nano to exist. Not to mention most full size cards can fit in micro ATX cases anyway.

THey are niche and under supported by manufacturers and underrepresented in the market RIGHT NOW, if the support for them improved and you could get the performance and price point in ITX that you currently do in mATX would the market shift? I think so, who wouldn't want to be able to have a mega gaming machine they could mount to the back of their TV via VESA and be done? Or stack a few up, a la linksys wrt54g routers, and fit everyone in the house's PC's on a shelf or in a closet and run Steam In-Home, shadowplay, or any other remote access or long run HDMI/USB and have your cake AND a clean client-side setup to boot?

 

ITX is a small market that shows the capability of going larger, Shield Consoles or SteamBoxes with full high end desktop GPU's while shrinking smaller and smaller, true portable gaming machines, or maybe another stab at smart modules for TVs, a standard mount and connection for an ITX box on the back of your favorite 60" OLED 4K 240Hz TV that turns it into an integrated PC-Smart TV. Its a full fledged Gaming monster run through your TV, with the capability to run your OTHER inputs through the PC. Console, stream or record gameplay, Cable or Satellite, Watch TV while doing your raids, Blu-Ray player, watch Das Boot while playing World of Warships or the World of Warcraft movie while grinding your PvP twink rogue. Who knows. 

[mrthuvi]

ITX systems are such a small niche part of the market their is absolutely zero reason for the Nano to exist. Not to mention most full size cards can fit in micro ATX cases anyway.

 

They do it because they can. Isn't it the point of technology? Sure, it is a niche market and it is priced as such. That much power in that tiny form factor didn't appear before and couldn't have appeared without HBM.

 

When HBM2 introduces, Nvidia will produce their own variant of the Nano anyway.

 

The future is already here and people don't realize it yet.

[zMeul]

I feel obliged to resurrect this topic over the debate me and some other had over the R9 Nano's thermal throttling - it thermal throttles! thanks @LinusTech for his latest video, watercooling a bunch of R9 Nanos:

 

 

my initial observation over the R9 Nano clocks was absolutely correct! the stock air cooling forces the card to thermal throttle; against many others saying, including Ryan Shrout, that it was reaching the power limit

 

#vindicated 

[ZetZet]

 

it's still reaching the power limit. you can't OC r9 nano for shit.