Probably the most anticipated card of the year due to its features, the Fury X was expected to be the slayer of Nvidia GTX 980 Ti but does it live up to the expectation?
No it doesn’t. The Fury X is one interesting card where the performance could be trailing the Nvidia GTX 980 one while outperforming GTX 980 Ti on another but we’ll cover that later. For a start, the card looks beautiful. It is short and compact, it doesn’t occupy many slots and it comes water-cooled.
The GPU section of the Fury X is sealed from every angle, even at the IO panel.
Here’s a view from the other side.
The Sapphire Fury X’s cables extends from the opposite side of IO panel instead of the top like the R9 295X2.
The radiator is cooled by a 120mm fan with curved blades.
Here’s the radiator for the Sapphire Fury X – it certainly looks familiar, so I did a comparison with the Cooler Master’s Nepton 120XL.
Oh hey, they look quite similar, don’t they?
Even from the side profile, it looks pretty much like the Nepton 120XL.
Here is the Sapphire Fury X along side the AMD R9 295X2 and a host of other AMD based graphic cards through the generations.
Starting from the top (the one with orange strip) – that’s the Sapphire Tri-X R9 390X, followed by the AMD R9 295X2. The 3rd one is of course the Fury X in this review, it is then followed by an AMD R9 270X reference design, an ASUS HD 7950 DirectCU II and lastly, a Sapphire R9 Dual-X 280X.
Phew, what a collection eh? 😀 Well, that’s goldfries’ lab for you.
Below is the GPU-Z capture for the card.
Did you notice that the bus width is 4096 bit? Check out the details of HBM (High Bandwidth Memory) at amd.com
More details of the Fury X at the official product page.
Test Setup
| Processor | Intel Core i5-4670K |
| RAM | Kingston HyperX 2x 4GB DDR3 2400Mhz Kit |
| Motherboard | Gigabyte GA-Z97X Gaming 3 |
| Cooler | NZXT Kraken X31 |
| Monitor | Dell U2414H |
| Power Supply | FSP AURUM S 600W |
| Casing | NZXT S340 |
| Operating System | Windows 8.1 Pro 64bit |
Benchmarks
Unigine Heaven 4.0
*NOTE : Details are set to maximum.
*NOTE : AA disabled in 4K test.
| Card | 1080p (avg) | 4K (avg) |
| Sapphire Fury X | 67.5 | 26.1 |
| ASUS GTX 980 Ti | 82.8 | 30 |
| GALAX GTX 980 SOC | 65.4 | 25.7 |
| AMD R9 295X2 | 97.6 | 33.5 |
Metro Last Light
| Card | 1080p (avg) | 4K (avg) |
| Sapphire Fury X | 99 | 38.5 |
| ASUS GTX 980 Ti | 90.5 | 40.5 |
| GALAX GTX 980 SOC | 93.5 | 35.5 |
| AMD R9 295X2 | 73 | 41.5 |
Bioshock Infinite
Settings are at
2 – UltraDX11_DDOF | 2 – Custom | 1 – 16:9 | 4 – FullHD / 4K
| Card | 1080p (avg) | 4K (avg) |
| Sapphire Fury X | 142.67 | 52.9 |
| GALAX GTX 980 SOC | 130.71 | 44.87 |
| AMD R9 295X2 | 134 | 68.57 |
Shadow of Mordor
Settings : Set to maximum, V-sync off.
| Card | 1080p (avg) | 4K (avg) |
| Sapphire Fury X | 94.11 | 36.03 |
| ASUS GTX 980 Ti | 98.74 | 35.01 |
| GALAX GTX 980 SOC | 76.56 | 29.03 |
| AMD R9 295X2 | 79.14 | 46.2 |
Grand Theft Auto V
Settings : Set to Very High for every possible option. 8xAA enabled for Full HD test, AA Disabled for 4K test.
| Card | 1080p (avg) | 4K (avg) |
| Sapphire Fury X | 48.36 | 40.77 |
| GALAX GTX 980 SOC | 62.14 | 48.29 |
| AMD R9 295X2 | 50.78 | 18.15 |
Temperature
Furmark Burn-in Test was used to stress the card. Fan settings are at Auto. Room set to ~25c.
| Card | Idle(°C) | Load (°C) |
| Sapphire Fury X | 31 | 61 |
| Sapphire Fury X Full RPM | 31 | 45 |
Furmark couldn’t detect the fan RPM and neither could Sapphire’s own TRIXX software. The consolation was that TRIXX’s fan control tab was able to control the fan. Judging by the sound, the fan seems to be running at around 50% when in AUTO mode.
The fan runs very silent at 50% but it’s the sound from the pump that does get a little irritating but it shouldn’t be an issue if it’s left in a covered casing.
Overclocking
The card cannot be overclocked. Running Sapphire TRIXX, the overclocking tab shows core and memory at 0 and the slider at the right most setting will show 10 on the interface.
Power Consumption
The stress was done with Furmark Burn-in Test. Power consumption reading was taken from the watt-meter, actual power draw by the entire system from the wall point. I’m using an FSP Aurum S 700W with 90% efficiency and the estimated system power draw (CPU, not including GPU) during Furmark test is 60w.
| Card | Sapphire Fury X |
| Furmark Burn-in | 336 |
| Estimated Actual System Draw | 302 |
| Estimated Card Power Draw | 242 |
The Verdict
The R9 Fury X retails at RM 2,899 – Not exactly an attractive price in my book but it’s a decent start. The reference model Nvidia GTX 980 cards are available at that price and offers better and more consistent performance in general. The Fury X lacks consistent performance, being better than GTX 980 Ti on some while trailing GTX 980 on others.
What you get with the Fury X is a card that’s different from the rest as it sports AMD’s High Bandwidth Memory and it comes water cooled which obviously is factored into the price. What’s even more amazing is that it is compact and it is practically not generating any heat from the GPU section of the unit. Overall I must say that the Fury X is something very interesting, and it certainly brought a fresh look upon an otherwise rather mundane scene of GPU warfare.
To make it simple, get the Nvidia GTX 980 Ti cards if you only care about performance. Get the Fury X if you want new technology on your system, one that looks cool, compact to fit even small casings and still holds the performance that hovers around the GTX 980 range.
The review sample was provided by IdealTech – if you’re looking for this card, you know where to find them.
