AMD’s R9 Nano Just Got More Interesting With Big Price Drop

Page 2 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.


If you wouldn't spend $500 on a graphics card, then why bother posting about that? You're completely ignore the merits of the product just because you aren't the targeted consumer for it.

The Radeon 7970 is about three years old and had a similarly high launch price. It had much more aggressive price drops due to competition once Kepler launched and today, it's still starting at about $150 on Ebay. The only options below that either not working or are auction listings that certainly won't end near $80. The Nano is a specialty card with little competition. I'd be surprised to see it selling under $200 in three years.

The 390 and 390X are too large and power-hungry to be usable in the situations the Nano is intended for, so the only real alternative is the GTX 970. Besides that, the 390 and the 390X require a more expensive power supply than Nano and the GTX 970, so a large portion of their price advantage is wasted. The few short 970 models are the only real alternatives for Nano and the performance difference is considerable.

Even with the 970 and ignoring the performance difference, you aren't saving "hundreds" (assuming you meant $ USD) because pcpartpicker shows us that the price difference for the cheapest models is about $170 (prices manually confirmed). The Nano is more expensive because it is not only just faster, but rather is the fasted card of its size available at this time. At $650, the price was too high to justify even that, but at $500, it is much more reasonable. This is especially true in resolutions above 1080p where the Nano creeps up on the now similarly priced 980.
 


390X even fits in the Raven, which is slightly bigger than an Xbox 1. I have mine in a 380T.
 

CarbonK

Reputable
Jan 12, 2016
11
0
4,510
Does anybody know how this card performs when liquid-cooled? I understand it's main disadvantage compared to the full-blown Fury X is the power circuitry, but maybe watercooling it can help get Fury X performance at some $150 less?

A few review sites have reported that an overclocked Nano with a stock air cooler is around 5-15% slower than a stock Fury X. Watercooling could probably close that gap to the point where the difference would be indistinguishable but buying a Nano waterblock would make the price difference indistinguishable as well.

The GTX 980 Ti still rules the absolute top end, but with the Nano at $500 and the Fury at $550 there really isn't a place for the GTX 980 anymore, unless you're looking for Nvidia-specific software compatibility.
 
Sep 30, 2013
281
0
10,810
You aren't considering power draw though.
The Nano's peak draw is 175w, which is a lot less than both of the options you just named.
It is still a niche product, but it caters to a group that very much exists.
TDP for GTX 980 185 watt.

So there's not much difference of the two one way or the other. The R9 390 is a different story though.
 

jerm1027

Distinguished
Apr 20, 2011
404
0
18,810


You aren't considering power draw though.
The Nano's peak draw is 175w, which is a lot less than both of the options you just named.
It is still a niche product, but it caters to a group that very much exists.

The R9 Nano's peak power draw is over 200W. TDP does not equal power draw. Watts is a measurement of energy, and the T in TDP stands for Thermal. Thermal Energy can, and is, measured in watts. Basically, the TDP is the heat output, and while that can give you a very rough idea of power consumption, it can't directly be compared to any non-identical GPU due to variances in efficiency of GPUs, even in GPUs of the same architecture (cut down GPUs tend to be less efficient than their full unlocked brethren). That means, hypothetically, a GPU1 could use more power than GPU2, but still have a lower TDP. Or vice versa - A GPU could use less power and still have a higher TDP compared to another GPU. Too add further complexity, especially if the CPU space, TDP is calculated differently from manufacturers, and a single TDP figure will be used across an entire product line, despite differences in power and performance and efficiency.
 

kcarbotte

Contributing Writer
Editor
Mar 24, 2015
1,995
2
11,785


That's true.
TDP is not peak power draw, and I knew that.

Doesn't change the fact that the peak power draw on the Nano is significantly lower than the Fury cards and lower tier AMD GPUs. Nvidia's 980 Ti isn't exactly economical either. It draws a fair bit of power, and outputs even more heat.

The main point is, there are customers for the Nano. They have very different priorities than the customers for the 980Ti and Fury cards. If you don't get what you'd want one, then you aren't the target customer.

 

JohnDR

Reputable
Jan 3, 2015
21
0
4,510
Just curious how the card performs when Bitcoin mining ? Not sure if I've seen too many articles talking about the HBM ram when Bitcoin mining.
 

kcarbotte

Contributing Writer
Editor
Mar 24, 2015
1,995
2
11,785


Bitcoin mining isn't very effective on GPUs anymore.
 

JohnDR

Reputable
Jan 3, 2015
21
0
4,510
Well see, this is a differently designed card, and from what I've read, it only consumes 175 Watts max while in use. So the question isn't, is it effective, the question is, "Is it efficient?" Cause for 175 watts, I could add four of those onto a MB and only draw 700 Watts, which is around what a Bitmain S5 runs, and I'm just curious to see if they would produce more than an S5 in efficiency. I see from the specs that that the Nano has 4096 shader processors. It would be interesting to see the output on a single Nano Fury when mining Bitcoin especially if was better than 2 Ghash per watt.
 
Status
Not open for further replies.