[nextpage title=”Introduction”]
We tested two of the most high-end video cards available today: the GeForce GTX 980 Ti from NVIDIA, and the Radeon R9 Fury X from AMD. Which one will deliver more performance in games? Check it out!
The GeForce GTX 980 Ti and the Radeon R9 Fury X are direct competitors, since they have the same price at about $650 USD. While the Radeon R9 Fury X (along with the R9 Nano) is now AMD’s most high-end video card, NVIDIA has the GeForce GTX Titan X (which we already tested) that targets the ultra-enthusiast market as it’s more expensive than the GeForce GTX 980 Ti.
You can see the tested video cards in Figures 1 and 2. They are both reference models from the chip manufacturers.
Figure 1: the GeForce GTX 980 Ti
Figure 2: the Radeon R9 Fury X
Figure 3 shows both video cards side by side, without their coolers.
Figure 3: both the video cards, without the coolers (Radeon R9 Fury X on top and the GeForce GTX 980 Ti on bottom)
In the table below, we compare the main specs from the video cards we included in this review. Prices were researched at Newegg.com for this article.
Video card |
Core clock |
Turbo clock |
Effective memory clock |
Memory bus |
Memory bandwidth |
Memory |
Processing cores |
TDP |
DirectX |
Price |
GeForce GTX 980 Ti |
1,000 MHz |
1,075 MHz |
7 GHz |
384 bit |
336 GiB/s |
6 GiB GDDR5 |
2,816 |
250 W |
12.1 |
USD 650 |
Radeon R9 Fury X |
1,050 MHz |
– |
1 GHz |
4.096 bit |
512 GiB/s |
4 GiB HBM |
4,096 |
275 W |
12 |
USD 650 |
Now let’s take a closer look to the tested video cards.
[nextpage title=”The GeForce GTX 980 Ti – part 1″]
In Figure 4, you can see the top of the GeForce GTX 980 Ti. It uses two slots, is 269 mm in length and weighs 923 g. The card uses two power connectors: one six-pin and one eight-pin. The green logo lights up when the card is powered.
Figure 4: top view of the GeForce GTX 980 Ti
Figure 5 shows the bottom of the card.
Figure 5: bottom view of the GeForce GTX 980 Ti
In Figure 6, you can see the video connectors of the GeForce GTX 980 Ti. It brings one DVI-I connector, one HDMI 2.0 output, and three DisplayPort 1.2 connectors. The card supports up to three simultaneous displays.
Figure 6: video connectors of the GeForce GTX 980 Ti
Figure 5 shows the solder side of the GeForce GTX 980 Ti. Notice that there are no memory chips and, just like on the GeForce GTX Titan X, there is no backplate.
Figure 7: solder side of the GeForce GTX 980 Ti
[nextpage title=”The GeForce GTX 980 Ti – part 2″]
Figure 8 shows the GeForce GTX 980 Ti with its cooler removed. The cooler is similar to the model we saw on the GeForce GTX Titan X. It touches the memory chips and the voltage regulator circuit transistors.
Figure 8: the GeForce GTX 980 Ti with the cooler removed
Figure 9 shows the GM200 chip. It is the same chip found on the Titan X, but with two SMM units disabled (the original chip has 24 SMM units), for a total of 2,816 cores, since there are 128 cores on each SMM. The rest of the chip specifications remain the same, like the 384-bit memory bus.
Figure 9: GM200 chip of the GeForce GTX 980 Ti
Figure 10 shows one of the 12 memory chips present on the GeForce GTX 980 Ti. It is a GDDR5 chip, model H5GQ4H24MFR-R2C from SKHynix, with 4 Gib (512 MiB) capacity and maximum clock of 3.5 GHz (7.0 GHz effective clock). So, there is no margin to overclock the video memory within its specs.
Figure 10: memory chip of the GeForce GTX 980 Ti
Figure 11 shows the voltage regulator circuit of the GeForce GTX 980 Ti. It uses six phases for the GPU and two phases for the memory chips, which is identical to the voltage regulator of the GeForce GTX Titan X.
Figure 11: voltage regulator circuit of the GeForce GTX 980 Ti
[nextpage title=”The Radeon R9 Fury X – part 1″]
You can see the Radeon R9 Fury X in Figure 12. The main visible feature of the card is its sealed liquid cooling system, mounted and installed; you just need to install the card in a PCI Express 3.0 x16 slow and attach the 120 mm radiator to the case. The card uses two eight-pin power connectors.
The card is 195 mm long, uses two slots, and weighs 880 g without the radiator (1,509 g with the radiator). The Radeon logo lights up when the card is powered up.
Figure 12: top view of the Radeon R9 Fury X
In Figure 13, you can see the video connectors of the Radeon R9 Fury X. It brings one HDMI (version not specified) output and three DisplayPort 1.2 connectors. The card supports up to six simultaneous displays (using a DisplayPort hub).
Figure 13: video connectors of the Radeon R9 Fury X
In Figure 14, you can see the back of the card. An interesting detail is the group of nine LEDs on the top left edge, which gives a real-time visual indication the of GPU usage.
Figure 14: back view of the Radeon R9 Fury X
Removing the back cover, you see the solder side of the Fury X. There are no memory chips here.
Figure 15: solder side of the Radeon R9 Fury X, without the cover
[nextpage title=”The Radeon R9 Fury X – part 2″]
Figure 16 shows the Radeon Fury X with its cooler removed. It is a liquid cooling block with copper base and embedded pump. The cooler structure is a big metal plate that keeps contact to the voltage regulator circuit transistors, and is connected to the block through a heatpipe. So, the liquid cooling system refrigerates not only the GPU, but all the “hot parts” of the video card.
Figure 16: the Radeon Fury X with the cooler removed.
In Figure 17, you can see the card. At first sight, it looks like there are no memory chips. Actually, the video card memory is located where the four small chips together to the main graphics chip.
Figure 17: the Radeon R9 Fury X without its cooler
Figure 18 shows the Fiji GPU. The biggest innovation on the R9 Fury X is the usage of HBM (High Bandwidth Memory). This memory uses a system where the memory chips are stacked and mounted on the same substrate as the graphics chip. With this system, it is possible to create a wider bus, when compared to traditional graphics card, where GDDR5 (or DDR3) memory chips are separated and connected to the GPU by PCB tracks. According to AMD, this standard also uses less power than the traditional DDR3 and GDDR5.
In the Fury X, the memory bus uses 4,096 bits. As this video memory works with 1 GHz clock, the available bandwidth is 512 GiB/s. The Radeon R9 Fury X comes with 4 GiB of HBM video memory.
The GPU has 4,096 processing cores, divided in 64 computing units.
Figure 18: graphics chip with the four HBM chips
Figure 19 shows the voltage regulator of the R9 Fury X, which uses six phases for the GPU.
Figure 19: voltage regulator circuit of the Radeon R9 Fury X
[nextpage title=”How We Tested”]
During our benchmarking sessions, we used the configuration listed below. Between the tests, the only variable component was the video card being tested.
On games, we ran the tests with both Full HD (1920 x 1080) and 4K UHD (3840 x 2160).
Hardware Configuration
- CPU: Core i7-5775C
- Motherboard: ASRock Z97 Extreme 4
- CPU Cooler: Intel stock
- Memory: 16 GiB DDR3-1866, two G.Skill Sniper F3-1866C10D-16GSR 8 GiB memory modules configured at 1,866 MHz dual channel
- Boot drive: Kingston HyperX Savage 480 GB
- Video Monitor: Samsung U28D590D
- Power Supply: Corsair CX750
Operating System Configuration
- Windows 10 Home 64-bit
- NTFS
- Desktop video resolution: 3840 x 2160 @ 60 Hz
Driver Versions
- AMD driver version: 15.7.1
- NVIDIA driver version: 358.50
Software Used
- 3DMark 1.5.915
- Battlefield 4
- Dirt Rally
- Dragon Age: Inquisition
- Dying Light
- GTA V
- Mad Max
- The Witcher 3: Wild Hunt
Error Margin
We adopted a 3% error margin; thus, differences below 3% cannot be considered relevant. In other words, products with a performance difference below 3% should be considered as having similar performance.
[nextpage title=”3DMark”]
3DMark is a program with a set of several 3D benchmarks. The Cloud Gate benchmark measures DirectX 10 performance, while Sky Diver also measures DirectX 11 performance being aimed at average computers. Fire Strike benchmark measures DirectX 11 performance and is targeted to high-end gaming computers. Fire Strike Ultra is the same test, but runs at 4K UHD resolution.
In this test, the GeForce GTX 980 Ti was 6.5% faster than the Radeon R9 Fury X.
On the Sky Diver benchmark, the GeForce GTX 980 Ti was 11% faster than the Radeon R9 Fury X.
On Fire Strike, the GeForce GTX 980 Ti was 5.5% faster than the Radeon R9 Fury X.
The Fire Strike Ultra is the “heaviest” benchmark available in this program. Here, both cards had similar performances.
[nextpage title=”Gaming Performance (part 1)”]
Battlefield 4
Battlefield 4 is one of the most popular games of the Battlefield franchise, being released in 2013. It is based on the Frostbite 3 engine, which is DirectX 11. In order to measure performance using this game, we walked our way through the first mission, measuring the number of frames per second (fps) three times using FRAPS. We ran this game at Full HD and 4K UHD, setting overall image quality at “ultra.”
The results below are expressed in fps and they are the mean between the three collected results.
On Battlefield 4, the GeForce GTX 980 Ti was 7.9% faster than the R9 Fury X at 4K UHD, and 21% faster at Full HD.
Dirt Rally
Dirt Rally is an off-road racing game released in April 2015, using Ego engine. To measure performance using this game, we ran the performance test included in the game, in Full HD and 4K UHD resolutions and image quality configured as “ultra” and MSAA 2x.
The results below are expressed in fps.
In this game, the Radeon R9 Fury X was faster than the GTX 980 Ti, outperforming it by 7.4% at 4K UHD and by 5.7% at Full HD.
Dragon Age: Inquisition
Dragon Age: Inquisition is the most recent game from the popular action RPG franchise Dragon Age. It was launched in November 2014 and uses the Frostbite 3 engine with SpeedTree.
We ran the game at 4K UHD and Full HD, with all quality options at “ultra,” measuring three times the frame rate with FRAPS. The results below are expressed in fps and represent the arithmetical mean of the three collected results.
On Dragon Age: Inquisition the GeForce GTX 980 Ti beat the R9 Fury X by 9% at 4K UHD and by 19% at Full HD.
[nextpage title=”Gaming Performance (part 2)”]
Dying Light
Dying Light is an open-world horror game launched in January 2015, using the Chrome Engine 6. We tested the performance at this game with all quality options at maximum and resolutions of 4K UHD and Full HD, measuring three times the frame rate using FRAPS.
The results below are expressed in fps and they are the mean between the three collected results.
On the Dying Light test at 4K UHD, the GeForce GTX 980 Ti was 11% faster than the R9 Fury X, and 47% faster at Full HD.
Grand Theft Auto V
Grand Theft Auto V, or simply GTA V, is a open-world action game released for PCs in April of 2015 and uses the RAGE engine. In order to measure the performance on this game, we ran the performance test of the game (the plane portion), measuring the frame rate with FRAPS. We ran GTA V at 4K UHD and Full HD, with image quality set as “very high” and MSAA as 2x.
The results below are expressed in frames per second.
On GTA V, the GeForce GTX 980 Ti was 13% faster than the Radeon R9 Fury X at 4K UHD. At Full HD, the difference was 18%.
Mad Max
Mad Max is an open-world action game launched in September of 2015, using the Avalanche engine. In order to measure the performance using this game, we ran its intro, measuring the framerate with FRAPS three times. We ran the game at 4K UHD and Full HD, with image quality set at the maximum.
The results below are expressed in fps and they are the mean between the three collected results.
On Mad Max, both cards had the same performance on both resolutions.
The Witcher 3: Wild Hunt
The Witcher 3: Wild Hunt is an open-world RPG released in May of 2015 and based on the REDengine 3 engine. In order to measure the performance on this game, we walked around at the first scene of the game, measuring the frame rate with FRAPS three times. We ran the game at 4K UHD and Full HD with image quality set to “ultra.”
The results below are expressed in fps and they are the mean between the three collected results.
On this game, both the video cards had similar performance in the 4K UHD test, but the GTX 980 Ti was 17% faster on the Full HD test.
[nextpage title=”Conclusions”]
Besides having very different technical specs, such as memory amount and type, core count, and even physical size, the video cards tested today are direct competitors since they have the same price.
While the GeForce GTX 980 Ti is a GeForce GTX TITAN X version with “only” 6 GiB of memory and 256 processing cores disabled, the Radeon Fury X brings innovative technology – the HBM memory – which allows this chip to be mounted on a PCB that is pretty smaller than the competitor.
In the matter of performance, however, they proved to be very close. While on the tests using Full HD resolution, the GeForce GTX 980 Ti had a small, but consistent advantage; using 4K UHD resolution, this advantage vanished. We can say that, on the games we tested using 4K UHD resolution, both the video cards have similar performances.
Another point that was clear is that both video cards are powerful enough to run the most recent games under 4K UHD resolution and maximum image quality with a frame rate above 30 fps. So, the gamer or enthusiast user who buys either one will be satisfied.
Leave a Reply