Logo   Banner   TopRight
TopUnder
Transparent
AMD Radeon HD 6970 & 6950 Debut: Enter Cayman
Transparent
Date: Dec 15, 2010
Section:Graphics/Sound
Author: Marco Chiappetta
Transparent
Introduction and Specifications
We have been hearing rumblings about the products we’ll be showing you today, which are based on AMD’s next-gen “Cayman” GPU, for months now. In fact, back in October we were able to leak a tiny bit of official information in our coverage of the AMD Radeon HD 6800 series, but there’s been rampant discussion, spy images, and other leaks from all around the web for ages. And why not? It has been almost 15 months since AMD launched the Radeon HD 5870, which remains the company’s fastest single-GPU. That’s a freakin’ eternity in Internet time.

As many of you may be aware, Cayman is the codename given to AMD’s newest flagship GPU, which will power the Radeon HD 6900 series (Antilles, aka the Radeon HD 6990, will uses a pair of Cayman GPUs when it arrives sometime next year). Cayman is the successor to the Cypress XT that was employed on the Radeon HD 5800 series, and while it does leverage many of the technologies from that generation, Cayman is a new chip that embraces a fresh core architecture designed to increase efficiency, performance and feature-set.

Cayman does not, however, employ a new manufacturing process. AMD’s typical M.O. is to release a chip using a given process, revise that chip at some point during its life-cycle to enhance performance and/or power, and then release a totally new chip built using a more advanced process 12 – 16 months later. This time around though, TSMC, AMD’s GPU foundry partner (for now) simply isn’t ready with their next-gen 28nm process, so Cayman, like Cypress, is built at 40nm. Because of this, Cayman is actually the largest graphics core that AMD has built to date.

Initially, the Cayman GPU will power two graphics cards, the Radeon HD 6970 and the Radeon HD 6950. And wouldn’t you know it; we have a pair of each in house and have been testing them with a completely revamped test-bed for a few days now. Take a moment to peruse the specs and soak up the first couple of pics of the new Radeon HD 6900 series cards below.  Then we’ll dive into the particulars and check out their performance.


Feast Your Eyes On The Radeon HD 6970
AMD Radeon HD 6970 and Radeon HD 6950
Specifications & Features



And Here's The Radeon HD 6950

As you can see in the two images above, the new Radeon HD 6970 and 6950 look virtually identical from the front, save for their model number decals. We suspect that those specifications may leave some of you scratching your heads, though. If the Radeon HD 6970 is supposed to be an upgrade over the 5870, why does it have a lower total number of stream processors? The Radeon HD 6970 has 1536 ALUs versus 1600 in the Radeon HD 5870, and the 6950 has “only” 1408 ALUs. Well, the answer to that question lies in AMD's new "VLIW4" architecture, which we’ll explain a little later.

Disregarding the shader counts for a moment, you can see the Radeon HD 6970 offers compute performance in the neighborhood of 2.7TFLOPs (single precisions) and 675GFLOPs (double-precision) and the Radeon HD 6950 offers 2.25TFLOPs / 564GFLOPs. The 6970's GPU is clocked at 880MHz and the 6950's at 800MHz. The Radon HD 6970 sports 96 texture units while the 6950 offers 88 (up from 80 in the 5870) and each has 32 ROPs. Both cards are also outfitted with 2GB of GDDR5 memory that utilize a 256-bit interface. The memory speed on the 6970 tops out at 5.5Gbps (1375MHz), while the 6950 hums along at 5.0Gbps (1250MHz). That puts memory bandwidth in the 175GB/s and 160GB/s ranges for the Radeon HD 6970 and 6850, respectively, which is a big step up from the 153GB/s of the 5870. Max board power is 250W for the 6970 and 200W for the 6950, and both have the same digital video output configuration.

Transparent
The Cayman GPU Architecture

At its core, the GPU codenamed “Cayman” is comprised of 2.64 billion transistors and its die size is roughly 389mm2. If you’re keeping track, Cypress (Radeon HD 5870) was comprised of about 2.1 billion transistors and had a 334mm2 die while NVIDIA’s GF110 is made up of approximately 3 billion transistors.  The actual die size for NVIDA’s part hasn’t been disclosed, however, numbers in the 529mm2 range have been thrown around. 


Cayman GPU High-Level Block Diagram

Here we have a high-level block diagram of the Radeon HD 6900 series GPU, aka Cayman. This chip is different from the previous generation Radeon HD 5800 series GPU in a number of ways. First and foremost, the 6900 series is outfitted with dual graphics engines which help keep the GPU well-fed with data, whereas the 5800 series had one. The 6900 series GPU also employs a VLIW4 (Very Long Instruction Word) design; the 5800 was a VLIW5 design. In its full configuration, the chip has 24 SIMD engines, 96 texture units, 32 ROPs (which have been upgraded from previous generations), a 256-bit memory interface and a number of enhancements to improve GPU compute features.


The Dual Graphics Engines Up Close

Cayman’s dual graphics engines give the chip the ability to process two primitives per clock and the dual-engine configuration offers double the transform and backface cullrate than the 5800 series. Incorporated into the engines are also the chip’s dual 8th generation tessellators which now have off-chip buffer support for high tessellation levels among other tweaks. AMD claims the new tessellators will perform up to 3x better than the 5870. Dual rasterizers allow the Radeon HD 6900 series GPU to process up to 32 pixels per clock.

  
The New Core (left) and Render Back End (right) Designs

The new VLIW4 cores have a 4-way co-issue design and all of the cores have equal capabilities. The VLIW5 cores in the 5800 series had a special “T-Unit” that was more capable than the other four units in the core. AMD claims the new core design offers a 10% improvement in performance per mm2 with simplified scheduling and register management.

The render back-ends, or ROPs, have also been upgraded in the 6900 series GPU. The new render back-ends in the chip allow for coalescing of write ops and they can now handle 16-bit integer (unorm/snorm) ops at twice the rate and 32-bit floating point (single/double component) ops are 2x to 4x faster. The end result is faster anti-aliasing performance that causes less of a performance hit than the previous generation.

Transparent
GPU Compute and New AA Modes

In addition to the new features we mentioned earlier, the Radeon HD 6900 series core offers a number of GPU-Compute and Image Quality related enhancements as well.


Cayman's GPU Compute Architecture

The Radeon HD 6900 series GPU’s compute enhancements include asynchronous dispatch support, which allows the chip to execute multiple compute kernels simultaneously with each kernel having access to its own command queue and protected virtual address domain. The GPU also sports dual bi-directional DMA engines for faster system memory reads and writes and the chip is capable of coalescing shader read ops.  It can also fetch direct to local data stores within on-chip memory with faster flow control and faster double precision ops, which are 1/4 of the single precision rate.


The Radeon HD 6900 Series' New EQAA Modes

New enhanced quality anti-aliasing modes, EQAA, are set to debut with the Radeon HD 6900 series as well. The new modes offer up to 16 coverage samples per pixel and the color and average samples can be independently controller. EQAA also offers custom sample patterns and filters, and the new modes are compatible with Adaptive, Super-Sample, and Morphological AA modes. EQAA should offer better image quality than standard MSAA modes (by all rights), but with a similar memory footprint requirement.

The 6900 series also has the new anti-aliasing mode that debuted with the Radeon HD 6800 series as well, Morphological AA. The Morphological AA feature is accelerated using DirectCompute and delivers full-scene antialiasing, but at speeds much faster than super-sampling. It is compatible with any DX9, 10, or 11 applications and is switchable via the Catalyst Control Center. Like the Radeon HD 6800 series, other enhancements to the Cayman / Radeon HD 6900 series GPU included a new Unified Video Decoder engine, bringing it up to UVD 3. The only major change here is that the new architecture can now accelerate DivX and xVid files as well. The Radeon HD 6900 series also features HDMI 1.4 and DisplayPort 1.2 support. They also support a new power clamping technology dubbed PowerTune.

PowerTune utilizes a control processor integrated into the Radeon HD 6900 series to monitor GPU activity in real-time and dynamically adjust clock speeds to enforce a hard TDP ceiling. PowerTune offers direct control over the GPU’s power draw and no longer needs to constrain default clock speeds to accommodate “power virus” type applications. We should also note that PowerTune is controllable via AMD’s Overdrive utility should users want to tweak performance based on their cooling and power configurations.

Transparent
The Radeon HD 6970 & 6950 Cards

The Radeon HD 6970 and 6950 are virtually indistinguishable from one another. The cards look identical, with the sole exceptions being the respective model number decals affixed to their fan shrouds and their power connector configurations.

  

  
The Reference Radeon HD 6970

Radeon HD 6900 series reference boards are 10.75-inches long and feature fan shrouds and heat plates that encase the entire card. On the back side of the cards, the GPU cooler retention bracket is exposed through a hole in the heatplate that runs the entire length of the card. The front side of the card is dominated by a black shroud with red accents that covers the card’s inner workings. Underneath the shroud lies a large vapor chamber cooler, with a heavy copper base and a dense array of aluminum fins. The barrel-type fan sucks air in through the far end of the card, where it is forced through the heatsink’s fins, and ultimately expelled out of the back of the system through vents in the case bracket. In singe-card configurations, we found the coolers on the Radeon HD 6900 series cards to be fairly quiet. At idle, the cards are essentially inaudible, masked by a system's CPU and PSU fans. Under load, the fans on the cards do spin up considerably and are audible, but they’re not very loud at all. In a dual card configuration, however, when the second card is butted up against the first, and the first card’s temps range considerably higher and as a result, its fan can spin much faster than normal and get relatively loud--definitely louder than the 5800 series.

   

   
The Radeon HD 6950 Reference Card

Along the bottom edge of the cards, the only thing visible is the PCI Express edge connector, but at the top there’s something interesting worth noting. Of course there are the CrossFire edge connectors for running cards in multi-GPU configurations, but nestled right alongside the second CF connector is a tiny switch. That tiny switch is used to toggle between two BIOS chips on the card—the first BIOS can be altered / updated, while the second will return the card to its factory settings. That’s something that might come in handy with the modding crowd.

Also on the top, at the far end of the cards are their power connectors. The Radeon HD 6970 requires two supplemental power feeds, a PCIe 8-pin feed and a 6-pin as well. The Radeon HD 6950 also requires dual supplemental feeds, but both of them are of the 6-pin variety.

The output configuration on reference Radeon HD 6900 series cards is unchanged from the 6800 series we showed you a few weeks back. They’ve got one singe and one dual-link DVI output, two mini-DisplayPort outputs and single HDMI outputs. Four of these outputs can be used at any given time to power displays in a multi-monitor Eyefinity configuration.

Transparent
Test Setup & Unigine Heaven v2.1

How We Configured Our Test Systems: We tested the graphics cards in this article on a Gigabyte GA-EX58-UD5 motherboard powered by a Core i7 980X six-core processor and 6GB of OCZ DDR3-1333 RAM. The first thing we did when configuring the test system was enter the system BIOS and set all values to their "optimized" or "high performance" default settings. Then we manually configured the memory timings (DDR3-1333, CAS 7) and disabled any integrated peripherals that wouldn't be put to use. The hard drive was then formatted and Windows 7 Ultimate x64 was installed. When the installation was complete we fully updated the OS and installed the latest DirectX redist, along with the necessary drivers, games, and benchmark applications.

HotHardware's Test System
Intel Core i7 Powered

Hardware Used:
Core i7 980X (3.3GHz)
Gigabyte EX58-UD5 (X58 Express)

Radeon HD 5870 (2)
Radeon HD 6870 (2)
Radeon HD 5970 (2)
Radeon HD 6950 (2)
Radeon HD 6970 (2)
GeForce GTX 570 (2)
GeForce GTX 580 (2)

6GB OCZ DDR3-1333
Western Digital Raptor 150GB
Integrated Audio
Integrated Network

Relevant Software:
Windows 7 Ultimate x64
DirectX Nov. 2010 Redist
ATI Catalyst v10.11b
NVIDIA GeForce Drivers 263.09

Benchmarks Used:

Unigine Heaven v2.1
Futuremark 3DMark11
FarCry 2
Just Cause 2
Alien vs. Predator
Metro 2033
Lost Planet 2
F1 2010*

* - Custom benchmark

Unigine Heaven v2.1 Benchmark
Pseudo-DirectX 11 Gaming


Unigine Heaven

The Unigine Heaven Benchmark v2.0 is built around the Unigine game engine. Unigine is a cross-platform real-time 3D engine, with support for DirectX 9, DirectX 10, DirectX 11 and OpenGL. The Heaven benchmark--when run in DX11 mode--also makes comprehensive use of tessellation technology and advanced SSAO (screen-space ambient occlusion), and it also features volumetric cumulonimbus clouds generated by a physically accurate algorithm and a dynamic sky with light scattering.

The new Radeon HD 6970 and 6950 are a huge step up over the Radeon HD 5870 in terms of performance according to the Unigine benchmark; the 6970 even manages to outpace the dual-GPU Radeon HD 5970 here. NVIDIA's GeForce GTX 500 series cards have an advantage, however, at the extreme settings we used to test. Geometry processing was a big focus for NVIDIA with Fermi and it shows in this test.

Transparent
Futuremark 3DMark11

Futuremark 3DMark11
Synthetic DirectX Gaming


Futuremark 3DMark11

The latest version of Futuremark's synthetic 3D gaming benchmark, 3DMark11, is specifically bound to Windows Vista and 7-based systems because it uses the advanced visual technologies that are only available with DirectX 11, which isn't available on previous versions of Windows.  3DMark11 isn't simply a port of 3DMark Vantage to DirectX 11, though.  With this latest version of the benchmark, Futuremark has incorporated four new graphics tests, a physics tests, and a new combined test.  We tested the graphics cards here with 3DMark11's Extreme preset option, which uses a resolution of 1920x1080 with 4x anti-aliasing and 16x anisotropic filtering.

3DMark11, which absolutely crushes all of the cards when tested using its Extreme settings option, tells an interesting story. Here, the Radeon HD 6970 pulls ahead of the GeForce GTX 570, but the 6950 trail's the 570 by a hair. When compared to the Radeon HD 5870 / 6870, the new Cayman based 6900 series offerings have a clear edge. That edge is gone when compared to the GeForce GTX 580, though. NVIDIA's current flagship is clearly the fastest single-GPU based card here. Nothing can touch the dual-GPU powered Radeon HD 5970, however.

Transparent
FarCry 2 Performance

FarCry 2
DirectX Gaming Performance


FarCry 2

Like the original, FarCry 2 is one of the more visually impressive games to be released on the PC to date. Courtesy of the Dunia game engine developed by Ubisoft, FarCry 2's game-play is enhanced by advanced environment physics, destructible terrain, high resolution textures, complex shaders, realistic dynamic lighting, and motion-captured animations.  We benchmarked the graphics cards in this article with a fully patched version of FarCry 2, using one of the built-in demo runs recorded in the "Ranch" map.  The test results shown here were run at various resolutions with 4X AA enabled.

From this point forward, we'll be comparing single-card and dual-card configurations in terms of performance, with the exception being the Radeon HD 5970 which gave us fits in a quad-CrossFire configuration this time around. Apologies for their omission from the multi-GPU graphs at the bottom of this and the processing pages.

As you can see above, in single card testing with FarCry 2, the Radeon HD 6900 series cards perform well. At the lower resolution they trail the GeForce GTX 570, but with the resolution increased to 2560x1600, the Radeon HD 6970 is able to match the performance of the GeForce GTX 570. The new Radeon HD 6900 series cards outpace the single-GPU 5870 throughout. The Radeon HD 5970 and Geforce GTX 580 are the clear leaders here.

Dual-card testing yielded some interesting results. Due to some excellent dual-GPU performance scaling, the Radeon HD 6900 series cards were able to take the two top spots here, outpacing even the GeForce GTX 580 SLI setup.

Transparent
Just Cause 2 Performance

Just Cause 2
DX10.1 Gaming Performance


Just Cause 2

Just Cause 2 was released in March 2010, from developers Avalanche Studios and Eidos Interactive. The game makes use of the Avalanche Engine 2.0, an updated version of the similarly named original. It is set on the fictional island of Panau in southeast Asia, and you play the role of Rico Rodriquez. We benchmarked the graphics cards in this article using one of the built-in demo runs called The Concrete Jungle.  The test results shown here were run at various resolutions and settings. This game also supports a few CUDA-enabled features, but they were left disabled to keep the playing field level. 

The Radeon HD 6900 series cards were significantly faster than the Radeon HD 5870 and 6870 in the Just Cause 2 benchmark at both resolutions, but they trailed the GeForces and Radeon HD 5970 by wide margins.

Once again, the Radeon HD 6900 series cards show good performance scaling in two-card configurations, but they weren't quite able to catch the GeForces.

Transparent
Metro 2033 Performance

Metro 2033
DirecX11 Gaming Performance


Metro 2033

Metro 2033 is your basic post-apocalyptic first person shooter game with a few rather unconventional twists. Unlike most FPS titles, there is no health meter to measure your level of ailment, but rather you’re left to deal with life, or lack there-of more akin to the real world with blood spatter on your visor and your heart rate and respiration level as indicators. The game is loosely based on a novel by Russian Author Dmitry Glukhovsky. Metro 2003 boasts some of the best 3D visuals on the PC platform currently including a DX11 rendering mode that makes use of advanced depth of field effects and character model tessellation for increased realism. This title also supports NVIDIA PhysX technology for impressive in-game physics effects. We tested the game resolutions of 1920x1200 and 2560x1600 with 4X anti-aliasing and in-game image quality options set to thei High Quality mode, with DOF effects disabled.


Metro 2033 shows us a lot of things. First, in this game the Radeon HD 6970 and 6950 end up being somewhat faster than the GeForce GTX 570. What's more interesting though is what happens to the cards equipped with 1GB frame buffers at the higher resolution. The Radeon HD 5870, 5970, and 6870 completely tank here when tested at 2560x1600, while the other cards are able to maintain decent framerates. In all fairness, this could be a drive bug, but it's possible the 1GB frame buffers aren't large enough to meet Metro 2033's needs at such a high resolution with AA enabled, whereas the 2GB Radeon card have some breathing room.

Once again, we see good performance scaling from the new Radeon HD 6900 series cards when running in a two-card CrossFire configuration and both the 6970 and 6950 are able to maintain their leads over the GeForce GTX 570 here.
Transparent
Lost Planet 2 Performance

Lost Planet 2
DirectX 11 Gaming Performance


Lost Planet 2

A follow-up to Capcom’s Lost Planet : Extreme Condition, Lost Planet 2 is a third person shooter that takes place again on E.D.N. III ten years after the story line of the first title. We ran the game’s DX11 mode which makes heavy use of DX11 Tessellation and Displacement mapping and soft shadows.  There are also areas of the game that make use of DX11 DirectCompute for things like wave simulation in areas with water.  This is one game engine that looks significantly different in DX11 mode when you compare certain environmental elements and character rendering in its DX9 mode versus DX11.  We used the Test B option built into the benchmark tool and with all graphics options set to their High Quality values.

Ouch. Lost Planet 2 is a hellacious benchmark that makes heavy use of tessellation. The new Radeon HD 6900 series cards are significantly faster than the Radeon HD 5870 and 6870 here, but they get trounced by the GeForces.

The thrashing continued in our dual card tests with Lost Planet 2. The GeForces simply dominate the Radeons in this benchmark.

Transparent
F1 2010 Performance

F1 2010
DirectX 11 Gaming Performance


F1 2010

Though Codemasters still continues to torture us with their ridiculously complicated labyrinth of game menus, we’ve found ourselves coming back to one of their titles for a taste of bleeding-edge DX11 benchmarking. F1 2010 is their latest racing simulation and like Dirt 2, it sports impressive visuals with DX11 support. “Ultra” settings for shadow effects and post processing elements like depth of field then become available to the gamer and in turn, crank up the workload on the graphics subsystem. The game engine also makes use of multi-core processors for higher performance on top-end systems. We tested the game configured with its Ultra graphics options at resolutions of 1920x1200 and 2560x1600.

The Radeon HD 6900 series cards performed well in the F1 2010 benchmark. Here, the Radeon HD 6970 finishes a step ahead of the GeForce GTX 570 and just misses the mark set by the GTX 580 at the higher resolution.The 6950 is faster than the GTX 570 at the higher resolution as well.

Again, due to some very good dual-card performance scaling, the Radeon HD 6900 series cards finish in the top two spots in the F1 2010 benchmark, besting even the GeForce GTX 580 SLI setup.

Transparent
Alien vs. Predator Performance

Alien vs. Predator
DirectX 11 Gaming Performance


Alien vs. Predator

The Alien vs. Predator benchmark makes use of the advanced Tessellation, screen space ambient occlusion and high-quality shadow features, available with DirectX 11. In addition to enabling all of the aforementioned DirectX 11 related features offered by this benchmark, we also switched on 4X anti-aliasing along with 16X anisotropic filtering to more heavily tax the graphics cards being tested.

Alien vs. Predator proved to be another strong point for the Radeon HD 6900 series. In this game, the Radeon HD 6970 and 6950 were able to outpace the GeForce GTX 570.

The same trend rang true in the dual-card tests--the Radeon HD 6970 and 6950 CrossFire rigs outpaced the GeForce GTX 570 SLI setup, but they couldn't quite catch the GeForce GTX 580.

Transparent
Total System Power Consumption

Before bringing this article to a close, we'd like to cover a few final data points--namely power consumption and noise. Throughout all of our benchmarking and testing, we monitored how much power our test system was consuming using a power meter. Our goal was to give you all an idea as to how much power each configuration used while idling and while under a heavy workload. Please keep in mind that we were testing total system power consumption at the outlet here, not just the power being drawn by the graphics cards alone.

Total System Power Consumption
Tested at the Outlet

Our power consumption tests don't show much to differentiate the Radeon HD 6970 from the GeForce GTX 570--both cards are within a few watts of one another under both idle and load conditions. Both cards also use much more power than the Radoen HD 5870 under load, which is an obvious result of their relatively large die sizes and high clock speeds. The Radeon HD 6950 also consumed a fair bit more power than the HD 5870, but the delta was much smaller.

With power consumption up from the previous generation, it should come as no surprise that heat output is increased as well. We found the Radeon HD 6900 series cards to idle in the low 40'C range but lead temps hover in the upper 80's.
Transparent
Our Summary and Conclusion

Performance Summary:  AMD's new Radeon HD 6900 series cards performed very well throughout our entire battery of benchmarks, but clearly pin-pointing their positions in the current 3D lanscape is somewhat difficult. We ran eight different games and applications to test the Radeon HD 6900 series cards. In four of those eight apps (3DMark11, AvP, F1 2010, and Metro 2033), the Radeon HD 6970 was faster than the somewhat similarly priced GeForce GTX 570.  In three apps it was slower (Unigine, Just Cause 2, and Lost Planet 2), and in the remaining one (FarCry 2) they performed at roughly the same level. We guess technically it would be fair to say the Radeon HD 6970 has a slight edge in overall performance, but in all fairness the trend could easily swing in the GeForce's favor with a different mix of apps that rely more heavily on tesselation effects. So, for all intents and purposes we'd have to call the battle between the Radeon HD 5970 and GeForce GTX 570 a draw, strictly in terms of performance. The Radeon HD 6950, however, has a clear edge over the Radeon HD 5870.  It's also going to debut at roughly the same price, and it offers 2x the frame buffer memory. If you were contemplating a Radeon HD 5870 purchase, scrap that idea and go for the 6950.


The Radeon HD 6970 and 6950, Side By Side

By now the reasoning for NVIDIA’s quick-fire, somewhat stealthy releases of the GeForce GTX 570 and GTX 580 should be clear. If AMD was able to get Cayman and the Radeon HD 6900 series cards out the door while the GeForce GTX 480 was still NVIDIA’s flagship GPU, AMD could have staked a claim for the fastest single-GPU and dual-GPU solutions on the market. As it stands today though, The GeForce GTX 570 definitely steals some of the Radeon HD 6900 series’ thunder and the GeForce GTX 580 remains the highest performing, single-GPU around.

Both the Radeon HD 6970 and Radeon HD 6950 should be hitting your favorite e-tailers immediately with MSRPs of $369 and $299, respectively. In light of our benchmark results, the Radeon HD 6970 may be priced just a tad high, but the more flexible output configuration with Eyefinity support and the large 2GB frame buffer warrant a small premium.

Now that we’ve seen what the Cayman GPU can do, we’re excited to see what AMD has in store with Antilles and the 6990. Looking at our power consumption numbers, it’s probably a safe to say that the Radeon HD 6990 won’t simply be two 6970’s bolted onto a single card, but more likely something along the lines of a hybrid mix of the 6970 and 6950, like the 5970 was in comparison to the 5870 and 5850. Regardless, looking at the dual-card Radeon HD 6950 CrossFire scores versus the single GeForce GTX 580 and the generally very good CrossFire scaling we saw throughout testing, we’re excited at the prospect of dual-Cayman powered graphics card. It may not be here in time for the holidays, but on the single-GPU front AMD is certainly competitive with NVIDIA again heading into the final days of the holiday buying season.

  • Strong Performance
  • Eyefinity Support
  • 2GB Frame Buffers
  • Relatively Quiet In Single Card Configurations
  • New AA Modes
  • Priced Competitively

 

  • 6970 Doesn't Clearly Beat The GTX 570
  • Can Be Somewhat Loud In Two-Card Config

 



Content Property of HotHardware.com