Sapphire AMD Fury Nano Review – Diminutive Size, Enormous Performance
The R9 Fury Nano from AMD is an interesting proposition, as it promises to give us near the full performance of their flagship R9 Fury X though at 64% decrease in power usage and in a compact enough package to be of use in builds that otherwise would have had to compromise on GPU power due to heat, power requirements or otherwise. The R9 Fury Nano, then, could very well have the potential to change how we look at and perceive the mini-ITX world.
Fury Nano: Sometimes great things come in small packages.
When Fiji XT was finally released to the world, it was an evolution of performance for AMD in a direction that has the potential to increase power efficiency across the board with the novel introduction of High Bandwidth Memory for the first time in any product. The very wide yet slightly slower memory bus is just one answer to the problem of a bottleneck that seems to inevitably exist in memory bandwidth. Not only that, but it shifted the focus back to a larger die and adding more resources as efficiently as possible. The catch is that it's still using a 28nm process, not exactly preferable, especially in the case of such a wee GPU as this.
|WCCFTech||Radeon R9 Fury X||Radeon R9 Nano||Radeon R9 Fury||Radeon R9 290X|
|GPU||Fiji XT||Fiji XT||Fiji Pro||Hawaii XT|
|GCN Compute Units||64||64||56||44|
|Render Output Units||64||64||64||64|
|Texture Mapping Units||256||256||224||176|
|GPU Frequency||1050Mhz||1000MHz||1000 MHz||1000Mhz|
|Memory||4GB HBM||4GB HBM||4GB HBM||4GB GDDR5|
|Effective Memory Speed||1Gbps||1Gbps||1Gbps||5Gbps|
|Cooling||Liquid, 120mm Radiator||Air, Single Axial Fan||Air, Custom AIB Solutions||Air, Single Blower Fan|
|Performance (SPFP)||8.6 TFLOPS||8.19 TFLOPS||7.2 TFLOPS||5.6 TFLOPS|
|Power Connectors||Dual 8-Pin||8-Pin||Dual 8-Pin||6+8 Pin|
|Launch Date||June 24 2015||September 2015||July 10 2015||October 24 2013|
The AMD R9 Nano is the answer to a question that likely hasn't really been answered before. One that has never truly been viable on quite this level. A full fat Fiji XT with all 64 compute units, 256 texture mapping units and 64 stream processors per compute unit for a total of 4096 Stream Processors. The same chip that's included in a water-cooled only Fury X design is placed on a 6" long PCB. The GPU has 1 HDMI 1.4a port and 3 DisplayPort connectors and a single 8-pin power connector.
The Nano is throttled, though at 85C, which is never really met unless under the most extreme of circumstances. Furmark is obviously one such example, and is certainly not representative of the real-world. Nominally, it stays well below that point. The power tuning software is also very much aware of the applications your using, moving up and down that power-curve depending in the demand placed on it, adapting pretty readily. The new power tuning algorithm by AMD should allow for a much more efficient GPU overall.
On a realistic level, the highest performance cards are limited by three factors from being viable in a Mini-ITX solution; availability of power-supplies to power such cards, thermals and size. A GTX T980 Ti isn't necessarily power hungry by any means, and could pair nicely with Silverstones newest 600W mini-ITX power-supply. The problem, however, is that it's stock cooling solution may not be adequate enough in small enclosures and it certainly is far too large to fit into the smallest of solutions. Until only recently, we weren't able to pair it with the highest end of processors anyhow, until ASRock's introduction of their X99 Mini-ITX board. Again, powering both a 16-thread processor and a high-powered graphics card as well as dealing with the head associated could be potentially threatening to the ultimate in performance.
AMD is positioning this as an ITX card, one that can physically fit in any case that allows for a full-height expansion card on the market. The idea is to provide a niche card to a niche market of builders and enthusiasts who want to build the most powerful, yet the smallest PC rigs possible. And this card is indeed quite wee.
The interesting thing is that while the R9 Fury X remains mostly unavailable due to the recent court ruling in the US regarding Asetek and Cooler Master, the R9 Fury Nano is sort taking up that price point as stock is available across the Internet. That also means that we don't have a sample of the Fury X to test against, because we simply can't get one without paying far more than retail.
The Sapphire R9 Nano follows the same standard OEM design created by AMD. It has a few novel features that should help to improve air-flow and thus keep temperatures at an acceptable level for a warmer than average chip. The fins of the heatsink are actually oriented horizontally, this way hot air is actually exhausted out the rear even for having an axial fan. AMD has also seen fit to include a dedicated heatpipe for the VRM, something they're saying is an industry first for an OEM design. On the underside of the actual heatsink itself you'll find a hybrid vapor chamber that includes flattened heatpipes that cover nearly the entire surface, aside from a vapor chamber that's positioned over the ASIC itself. All of those changes add up to a something of a miracle in the cooling realm. As we'll see later, those minor changes make a huge difference in actual temperature of the card while gaming.
The R9 Fury Nano is clocked fully at 1000MHz in most situations, but has a much more dynamic clock-speed that allows it to adjust to a certain power target and temperature target on a far more incremental scale. Generally we'll see clock-speeds between 800MHz and 950MHz to support the power target.
Sapphire R9 Nano Specifications
|Sapphire R9 Fury Nano Specifications|
|Chipset||Radeon R9 NANO|
|Core Clock||1000 MHz|
|Memory Clock||1000 MHz|
|Process Technology||28 nm|
|Memory Size||4 GB|
|Memory Bus||4096 bit|
|Card Bus||PCI-E 3.0|
|Digital max resolution||4096 X 2160|
|Analog max resolution||2048 x 1536|
|I/O||HDMI / Display Port*3|
|Power requirement||600W (with one 8-pin external power connector)|
The question that should be answered first is whether or not the Nano can actually meet it's power targets at stock settings, and keep things within an acceptable temperature range. This is marketed towards the Mini-ITX, so why don't we throw this into a very small case to see how it can handle dissipating heat. The power tests were ran using the test rig listed below. But to truly test the temperature, we put the R9 Fury Nano into the smallest case on hand, the Cooler Master Elite 110. This was placed with an AMD Athlon X4 860K with the stock cooler, not a cool running chip. The rest of the cards were on the normal test bench, because they wouldn't fit in the Elite 110.
This was designed to test the notion that the Nano could do well in enclosed spaces that may not necessarily have the best engineered ventilation. To be sure, this is good news because if it can maintain a reasonable temperature in a small space, then it will surely do well in a more well ventilated situation. In fact, in the open-air test bench, temperatures remained at 30C at idle and were 72C at load using the same Crysis 3 test.
But what of power? Switching over to the test bench, how does the whole system do with the Nano installed? The new power tests are conducted using a normal gaming load, so as to show a more average power draw that you can expect. This is full system power draw as measured from the wall. Idle power draw was measured after 15 minutes of inactivity.
All well and good, a total power consumption of 320, less then that of the vanilla Fury and certainly less than the Titan X or 980 Ti. The power consumption looks fairly good, certainly moving more toward the ideal for the Mini-ITX enthusiast or builder. Thus far, my example, a retail sample, doesn't exhibit any sort of mechanical or electrical noise that I can tell. So what of the noise levels?
Noise levels were measured using a Triplett Mini Sound Level Meter pointed at an open case 100cm away. This measures entire system noise from an open test-bench.
|CPU||Intel Core i5-6600K|
|Motherboard||ASRock Z170 Extreme 4|
|Power Supply||EVGA SuperNOVA 1300 G2|
|HDD||SanDisk Extreme II 120GB|
|Storage Disk||Seagate 2TB|
|Memory||16GB Crucial Ballistix DDR4 2400|
|Video Cards||AMD R9 Fury, AMD R9 Fury Nano, GeForce GTX Titan X|
|Operating System||Window 10 64-Bit|
For all the tests MSAA was set to X2 to even the playing field. Battlefield 4 consisted of a play through of a 64 player server on the Siege of Shanghai level. Crysis 3’s benchmark was done by playing through the first level. Fraps was used to capture the framerates of Battlefield 4 and Crysis 3. The rest use the internal benchmarking services that were available.
Because the R9 Fury Nano is being touted as being able to enable the 4K experience in small spaces, we'll look at 4K first to determine just precisely how well it can handle such a resolution against its similarly priced rivals.
Unfortunately, I don't have the closest rival on hand in terms of size. Asus's GTX 970 Mini-ITX would match the size, though it isn't nearly in the same price range nor does it, as you'll see, even perform nearly as well as the Nano does. The R9 Fury Nano is literally in a class of it's own, though perhaps the price of entry might be a bit too steep for some.
The Frostbite engine has come a long way in the many years of its development. Even though it was released in 2013, it provides great insight into the capabilities of the modern GPU. DICE has done a great job making it look fantastic while generally running well on any number of GPU's It's still quite the challenge for even the most potent, however.
Crysis 3 is still one of the best looking games around. It's a challenge for any video card and almost the standard for which we jokingly ask if something is powerful enough. The CryEngine is a very advanced engine and brings most GPU's to it's knees when everything is turned on, such as here.
Dragon Age: Inquisition
Released in in 2014, Dragon Age: Inquisition makes use of the Frostbite 3 engine in a slightly different manner and perspective. It also represents one of the better looking RPG's out there. Being Frostbite based, it should prove challenging to most GPU's.
Middle Earth: Shadow of Mordor
Shadow of Mordor is the wild card here, and I didn’t really know what to expect from it. The modified LithTech engine is an almost unknown. It has the ability to provide some nice looking visuals in a visceral world within Lord of the Rings, with great lighting and particle effects.
Civilization: Beyond Earth
As a 4K turn-based strategy game, one wouldn't necessarily think that this is a tremendously graphical intensive game, but it is. It makes heavy use of shader effects to make the map look and feel alive. That, and having the entire gigantic map filled with units does take its toll on the GPU.
Grand Theft Auto V
GTA V is a gorgeously rendered game, and very memory intensive at that. It even has a built in memory usage meter so you can see precisely what effects use the most of your precious graphics VRAM. That being said, there are a tremendous amount of options available in game to mess with, and doesn't have any pre-defined graphical levels. High Quality here means everything is turned up to it's highest, we have the softest shadow setting and MSAA is set to 2X on reflections and in the game itself.
The Witcher 3: Wild Hunt
This is perhaps one of the most demanding games in our test arsenal. It makes use of some very high resolution textures and nearly every effect under the sun that you could possible think of to provide a very emotional and engrossing experience. It's a fantastic looking game and could very well be the new standard for GPU performance. This was benchmarked using version 1.5.
Far Cry 4
The open-world and lush nature of Far Cry 4 means that it's a very demanding game. This does make use of NVIDIA's GameWorks, though even when turned off the playing field is relatively level, and it is a still a very tough test for both AMD and NVIDIA cards. This runs on the Dunia 2 engine, which has a very well done global illumination system, a realistic weather system and the capability of rendering some very dense and graphically taxing vegetation.
Assassin's Creed Unity
Unity is being included in the benchmark suite due to it being seen as a poorly optimized game for both brands despite being a GamesWorks title. AnvilNext 2.0 is capable of scaling very well, being in For Honor, Rainbow Six: Siege Ghost Recon Wildlands and of course the upcoming Assassin's Creed Syndicate. AnvilNext can potentially provide gorgeous visuals with lush vegetation, realistic character animations and support for a great in-engine physics API. This will also help to provide a basis for gauging the improvement of the engine over time as newer titles with it are released. So for now, just enjoy the low FPS.
The results, when taken merely as a price/performance ratio, are a mixed bag. On one hand we see that it's 90% the performance of the vanilla Fury even though it has more CU's and operates at a marginally lower frequency. The performance is not necessarily worth the money based off of that alone. On the other hand, we have a price/performance ratio of a card that has no equal in terms of its size and stature. While it's cool, quiet and it performs well, it isn't going to be at the top of most peoples list simply do to the niche market it wishes to carve.
Fourteen Forty Pea
Next up we'll test the more pedestrian and more available resolution of 2560x1440P, which is becoming more and more affordable as well as evolving with faster panels that support both FreeSync and G-Sync.
The results are a bit as we would expect, again nearly 90% or more of the Fury itself. the FPS is consistently higher than the 980, though it does fall short of the 980 Ti and of course the more expensive Titan X. For the segment they're wishing to enter, however, that's better than any card in it's class.
And of course, below is a compilation of games to show off how the Fury Nano does in games that are either older, or perhaps not quite the most popular releases at the moment. All of the games below were run at 1440P with the highest settings possible.
The R9 Fury Nano is really an anomaly. It's at once a unique and powerful product, capable of good performance at 4K and great performance at 1440P for it's size and the amount of power it draws. The temperatures it maintains also speak to some very imaginative engineering on the part of AMD. But it's also a niche card that is priced higher than peers that obviously perform better than it does in the same situations. The Nano is a good card, a great card and an especially beautiful card, but the price of it means that you're paying for the over-engineering to be able to stuff this into a far smaller space than you would anything else that performs like it.
And that's where the Nano does stand out, though. That small dominion that it rules over, it absolutely dominates. It's the highest performing card in it's class, being of the Mini-ITX variety. It has no rival for the very specific market that it fits within.
So for those looking for the absolute best 4K experience that money can buy, this isn't it. But if you're looking for a pairing that provides the smallest package of power for the money, then this is most definitely it. So no, it's likely not for you if you aren't looking to build the smallest yet most powerful system ever. Just imagine stuffing the Xeon E5-2687W V3 with 10 Haswell cores clocked at 3.1GHz on an ASRock X99 Mini-ITX with an R9 Fury Nano inside something the size of the Cooler Master Elite 110 or smaller. Now that would be a very mighty system indeed.
Just keep in mind that this isnt the be-all-end-all for performance, but it is for Mini-ITX performance.