AMD Radeon HD 6870 & HD 6850 Review

[SKYMTL]

Back on September 23rd 2009 the world was introduced to the HD 5870 and HD 5850; ATI’s first entries into what would quickly become a very lucrative DX11 marketplace. These graphics cards (code named Cypress) belonged to the Evergreen family and have enjoyed a huge amount of success. Naturally, part of that success was due to NVIDIA’s oft-delayed Fermi products but even with the GTX 400 series making some headway, ATI’s own DX11 parts are still selling like hotcakes.

Even though forward movement of in-game graphics technology seems to have nearly ground to a halt due to the extended life cycles of current consoles, AMD is now refreshing their mid and high-end lineups with a whole new range of GPUs. Code named Northern Islands, this family of graphics cards is a simple evolutionary step for the HD 5000 series, and should set a new benchmark for the price you pay for performance. This evolution can be likened to Intel’s tick / tock mentality where new chip designs are released every other cycle while refinements are rolled out in a minor “refresh” at the mid way point of an architecture’s lifespan. AMD has rolled a number of efficiency, tessellation and other improvements into their HD 5000-series in order to bring the HD 6000 series to market. They also happen to be the first products which ditch the “ATI” moniker.

The first Northern Islands products out of the gate will be the HD 6870 and HD 6850 – code named Barts XT and Barts Pro. Instead of going with the usual high-end first mentality, AMD is gearing the 6800 series towards the $199 to $250 market. According to our contacts at board partners and retailers, we should see the HD 6850 hit shelves for between $179 and $199 while the HD 6870 will go for $239 to $249 USD. These prices should make it quite obvious that Barts aims to attack the GTX 460 768MB and 1GB products head on. The result could be a quick marginalization of two products which NVIDIA has been banking on for the last few months.

The HD 5000 cards have proven their staying power as evidenced by AMD’s refusal to issue any meaningful price cuts even with mounting NVIDIA pressure. Nonetheless, we’re about to see if AMD has been able to regain their lead against NVIDIA’s latest and greatest.

 

[SKYMTL]

Diving into the Barts Core

Diving into the Barts Core

Unlike what they did with the Evergreen series, AMD isn’t trying to rewrite the book on performance or push new boundaries with their “refreshed” cards. Rather, the dual engine architecture which distinguished Cypress has been generally carried over with a few modifications made along the way. Barts isn’t the focus of a fundamental architectural change in any way, shape or form. It is all about the gradual refinement of an existing design into something with with a smaller die size and superior performance per watt. The end result is that AMD can now push a more affordable high performance solution to end users without sacrificing profit margins.

Cypress Core​

At first glance, there isn’t all that much to distinguish the full Barts XT core from the outgoing Cypress other than the obvious change in the number of SIMDs, which results in less overall SPs (or Stream Cores as AMD calls them). In order to achieve high end performance which is optimized for efficiency, the engineers started with the basic back-end of the Cypress XT and built up from there. This means the graphics engine including the fixed function stages, L2 cache, ROP arrays and memory controllers have gone largely untouched. There were some changes to improve tessellation performance and communication between the different stages within the rendering pipeline but the vast majority of tweaks happened within the SIMD engine layout.

Since the Cypress Pro (which AMD is replacing with the Barts XT) used a full Cypress XT core with a few disabled SIMDs, it was inherently inefficient from a number of perspectives. In order to increase performance per watt, the HD 5850 was taken as a benchmark and the engineers set about trying to match the “sweet spot” it occupied in the market with a slightly revised layout.

The Barts core in both its XT and Pro forms retains the same 80 SPs along with four texture units, 32KB of Local Data Share and 16KB of L1 texture cache per SIMD as the Cypress series. However, where things have changed lies in the total quantity of SIMDs per core which has shrunk from 20 down to the 14 we now see in Barts. This in effect lowers the maximum possible SPs from 1600 down to 1120 and the number of TMUs from 80 to 56. However, since the render back-ends aren’t touched, the Barts XT has a full 32 ROPs. The memory interface also remains at 256-bit for the GDDR5 which is actually a first for an architecture that is aimed exclusively at the sub-$250 market.

From these descriptions and the fact the parts carry the HD 6800 moniker you may assume that Barts is meant to be a direct top to bottom replacement of the HD 5870 and HD 5850, but it isn’t. There will be some overlap, but for the time being the Barts XT -or HD 6870- will take up the HD 5850’s mantle while the HD 5870 will be pushed aside by a completely different beast.

Other than the obvious changes to the SIMD layout, there has also been some window dressing going on behind the scenes. The main graphics engine which entails the fixed function stages of AMD’s architecture is for the most part carried over from the HD 5800 series without any significant changes but there is one major addition: an enhanced tessellator.

One of the main critiques leveled against Cypress series GPUs was their tendency to choke under heavy tessellation workloads. Through improved thread management in the shader engines as well as enhanced buffering for tessellation draw calls, AMD has been able to manage up to a twofold increase in overall tessellation performance over the HD 5800 cards. We can also see that in an effort to increase rendering efficiency even more, AMD has broken up the Ultra Threaded Dispatch Processor into two with each section having its own instruction and constant cache. This dispatch processor basically acts like a traffic cop, directing draw calls to the SIMD arrays. With each directing its own “half” of the SIMD engine, rendering information can be processed at a much quicker rate without adding to the overall die size of the Barts.

To put this into layman’s terms, the Barts architecture is able to remove the tessellation bottleneck which allows more of the rasterizers and SPs to be used more efficiently and as a result DX11 performance in particular has been increased.

Basically, the architectural tweaks AMD has made are mainly focused upon improving DX11 and tessellation rendering efficiency but in doing so there have been a number of tertiary benefits as well -such as a large increase in geometry performance.

With the introduction of the Barts core, AMD has taken the first step towards what they call the balancing and evolution of the HD 5000 series architecture. At 1.7 billion transistors Barts is actually smaller than Cypress and uses roughly 25% less silicon yet can achieve better performance than the HD 5850. Naturally, most of the performance per area increases have been attained through higher core and memory clocks but the added efficiency brought about by less SIMDs and streamlined communication between the numerous stages in the rendering pipeline have a significant effect as well. AMD hopes that all of these minor changes have the ability to augment performance to better compete with NVIDIA’s current offerings.

 

[SKYMTL]

AMD’s Current and Upcoming Lineup

AMD’s Current and Upcoming Lineup

With the addition of the HD 6800-series into AMD’s lineup, there are bound to be some toes stepped on in the name of progress. The Nothern Islands GPUs will eventually take over the entire top to bottom product stack but for the time being the HD 6870 will replace the HD 5850 while the HD 6850 will take over from the HD 5830.

As we discussed on the last page, even though the Cypress cards have the HD 6800 series beat in terms of SPs and texture units, the new cards have a number of optimizations built into their architecture which allow for higher performance per square millimeter. Another benefit of these smaller cores is their ability to pack more performance into less transistors which leads to an increase in performance per watt as well. Even the HD 6870 which operates at a blistering 900Mhz core / 1050Mhz (4.2Gbps) for memory clocks consumes a maximum of 151W.

When it comes to the actual cost of these new graphics cards, you can expect them to be priced according to their relative performance against the HD 5800 series as well as the competition’s products. That being said, we expect to see the HD 6870 going for around $250 USD while the HD 6850 should retail for somewhere in the neighborhood of $179 - $199.

The entire point on this exercise in affordable prices is to replicate the pricing structure of the HD 4800 series; in particular the HD 4980’s initial SRP of $230. This hit what AMD calls the “sweet spot” in the market and we have to agree with them. The HD 5770 toyed with this $100 gap when it was introduced at $160 and the HD 5850 narrowly missed it by a good $50 upon release. Ever since the HD 4800-series cards there really hasn’t been much ATI / AMD competition against NVIDIA between the $150 and $250 price points other than the HD 5830. Obviously, that’s about to change.

So why did AMD name these cards the HD 6800-series if the higher end Barts XT has its paper specifications manhandled by the HD 5850? For starters, the HD 5700 series of Juniper-based cards will be staying with us until at least Q1 2011, at which point they will likely be replaced if AMD sees the need. Allowing the Barts cards to carry the HD 6700 moniker while being sold alongside the HD 5770 and HD 5750 would have caused more confusion in what is already a market with far too many overlapping products. This move has also opened up AMD's lineup for additional SKUs in markets that have been historically ignored.

Believe it or not, Barts only represents a small piece of AMD’s pie over the next quarter or so since they will be introducing a whole range of products that cover literally every price point from $199 on up. November will see the release of the Cayman XT-based HD 6970 and Cayman Pro-based HD 6950 which have all of the features seen in Barts plus enhanced rendering scalability and off-chip buffering for DX11 applications. These will be the spiritual successors to the HD 5870 and HD 5850 and should go head to head with the higher end Fermi cards.

December will see the introduction of Antilles which is meant to be the lynchpin of AMD’s renewed assault on the DX11 market. The HD 6990 will bring untold performance to the table through the use of a pair of Cayman GPU cores and additional features we can’t divulge at this time.

Before the introduction of the HD 6870 (Barts XT) and the HD 6850 (Barts Pro), there was a yawning gap between AMD’s budget friendly HD 5770 and the decidedly higher performing HD 5850. This chasm was bridged for a time by the HD 5830 but it exhibited inconsistent performance and never really had a chance of competing against NVIDIA’s GTX 460. This resulted in the GF104 based card garnering a large following in a short time frame. Indeed, the entire reason behind AMD’s push to release the HD 6870 and HD 6850 prior to the higher end HD 6900 series is to quickly cut into NVIDIA’s dominance of this highly popular market.

 

[SKYMTL]

Tessellation Done Right?

Tessellation Done Right?

We all know that tessellation can be used by a developer to add a large amount of high detail geometry to a scene, but in the past this meant a hefty performance penalty for AMD’s DX11 cards. Without the ability to significantly change the way in which the architecture processes tessellation, AMD needed a way to decrease the overall “cost” of implementing one of the integral features of the DX11 ecosystem.

In the previous pages we mentioned that AMD has augmented tessellation performance on the Northern Island family of graphics cards and to do so they basically increased the efficiency at which tessellated scenes are processed.

AMD claims the current way of implementing tessellation is wasteful. According to a paper published at SIGGRAPH 2010 (PDF), the excessive use of tessellation can lead to sub-optimal rasterizer utilization and increasing the triangle mesh size to include up to 16 pixels wouldn’t decrease image quality but would vastly increase tessellation efficiency.

In addition to the under utilization of a GPU’s rasterizers, the issues with low pixel to triangle ratios have been far reaching. Due to the massive number of polygon edges, implementing MSAA algorithms can cause a staggering performance hit which runs counter to DX11’s claims of streamlining the rendering process. Developers have also run into problems with over-shadowing which again causes a performance hit and also reduces the overall realism of a given scene.

As the triangle size increases to encompass more pixels, the overall number of shader passes per pixel decreases as well but there is a trade off in terms of overall image quality and detail. In order to overcome this, developers have to strike a balance between shader / tessellation performance and overall mesh fidelity.

It is important to remember that these advances cannot be made at the hardware level; developers have to begin using the methods necessary to increase tessellation performance across a wide cross section of hardware. In these slides, AMD is only showing what is possible if developers use methods for higher tessellation efficiency in their upcoming titles.

While most developers have yet to begin utilizing different tessellation methods, AMD is moving forward with their own efficiency improvements for the Northern Islands architecture. This involves using some of the ideas we outlined above but transporting them to the hardware level instead of waiting for developers to play catch up.

Achieving higher tessellation efficiency for the HD 6000 series involves a number of hardware tweaks along with pushing a method called adaptive tessellation. Adaptive tessellation involves applying higher levels of tessellation to objects that are closer to the camera while objects further away will be rendered using lower levels. Using this type of method could also decrease the performance impact of applying certain anti aliasing algorithms to tessellated scenes.

Does this translate into higher performance in comparison to the HD 5000-series? Yes, but only at lower tessellation levels. However, once the tessellation factor increases beyond a certain point, the overall tessellation performance of AMD’s HD 6000 series levels off and is only slightly above a HD 5800 series card.

 

[SKYMTL]

Image Quality Improvements Aplenty

Image Quality Improvements Aplenty

Along with refining the architecture, AMD has also brought some new image quality enhancements to the table. The cornerstone of this push to increase IQ is the addition of a new anti-aliasing method which AMD calls Morphological AA.


Morphological AA Explained

Morphological AA is basically a new form of fullscreen anti-aliasing that delivers an image quality which is comparable to Super Sample AA, but can be implemented with a fraction of SSAA’s performance hit. The AA algorithms are calculated more efficiently by leveraging the GPGPU compute abilities of modern Radeon cards and the power of the DirectCompute API. Since the post-processing filtering is done by DirectCompute, the whole scene can be quickly analyzed so this AA method isn’t limited to only certain aspects of a given image.

One of the more interesting benefits of Morphological AA being done through a standalone API is the fact that it can be applied to both 2D and 3D scenes. It can be applied to things like video, Flash apps and more. In addition, since it is controlled directly through AMD’s Catalyst Control Center and makes use of DirectCompute, Morphological AA has the ability to be forced in any DX9, DX10 or DX11 game.


Anisotropic Filtering to the Next Level

On top of Morphological AA, AMD has also added new anisotropic filtering algorithms which are supposed to increase the accuracy of angle dependent AF in most situations.

In most situations, when you are blazing through a level you likely won’t take notice of properly executed anisotropic filtering but there are plenty of times when low quality AF algorithms stand out like a sore thumb. An example of this can be seen in Dragon Age: Origin’s many cobblestone passages: try to do things the quick and easy way and the cobblestones can degrade into a visual mess a few feet in front of the player.

What AMD is trying to accomplish is an improvement upon their already class-leading AF performance in order to address the discontinuity sometimes seen in very noisy textures. In order to do this, smoother transitions between filtering levels have been implemented so the sometimes jarring change between textures in the distance isn’t as apparent. The resulting image quality increase should be relatively minor but supposedly this new implementation won’t cause any type of performance hit when compared to other AF methods.

To consolidate all of these new image quality enhancements, AMD has rationalized the Catalyst Control Center so they can be modified through the use of a simple slider and check box system.

 

[SKYMTL]

AMD’s Eyespeed & UVD3

AMD EyeSpeed

The Northern Islands family of GPUs has a whole stable of video playback features, and in order to keep them all under banner AMD created the Eyespeed brand. Eyespeed will be the all-encompassing term given to all of the multimedia enhancing technologies such as pre / post processing, transcoding and HD playback within a unified ecosystem.

Eyespeed is split within two main spheres of influence: leveraging parallel processing for improved performance and video decoding through AMD’s UVD3.

UVD3; UVD2 on Steroids

As many of you probably already know, AMD’s Universal Video Decoder has been around for years and is known as one of the most capable video processing platforms currently available. UVD is now taking the next logical step forward with an expanded list of accelerated codecs in addition to the ones which were already compatible with past iterations.

One of the main features which have been added to the newly minted UVD3 is the ability to decode videos which use MVC encoding. As part of the H264 / MPEG-4 AVC codec, MVC is responsible for creating the dual video bitstreams which are essential for stereoscopic 3D output. Supporting this standard brings AMD’s GPUs the ability to process Blu Ray 3D movies through a HDMI 1.4a connector.

MPEG-4 Part 2 hardware acceleration for DivX and Xvid codecs has also been added but there is no mention made about the Nero Digital codec. For the time being, we’ll assume that Nero Digital acceleration will be added at a later date.

AMD’s main focus for these new graphics cards and for future products is to go beyond simply decoding HD content and instead add high end image quality improvements prior to the signal reaching the display. Through the use of the compute resources within a given system, additional pre and post-processing can be done before outputting an HD video stream.

AMD claims the additional processing their products can accomplish will significantly boost overall image quality whether it is for a simple upscale standard definition image or a true high definition video stream. The HQV benchmark is a highly, highly, highly subjective metric in which to determine image quality but supposedly the HD 6000 series will simply destroy the competition once all of its video processing features are enabled.

 

[SKYMTL]

HD3D: AMD Jumps into the Stereo 3D Pool

HD3D: AMD Jumps into the Stereo 3D Pool

Stereoscopic 3D has been the talk of the town for the last few years and AMD is understandably anxious to get in on the game. Unfortunately, quite a few consumers have been turned off of 3D movies due to studios ramming them down our throats even though many fail to live up to the benchmark set by Avatar.

The Stereoscopic gaming market on the other hand has been dominated by NVIDIA’s 3D Vision which we have complimented again and again. It offers excellent driver support and through 3D Vision Play can be used on a large number of LCD / LED TV screens as well. AMD on the other hand is promoting their “HD3D” as an open standard that is compatible with a number of third party stereo 3D driver wrappers, active shutter glasses and monitors. And yes, just like NVIDIA, AMD has chosen to go the active shutter glasses route.

In order to build a structure in which stereo 3D games can be developed and 3D movies played on AMD’s hardware, a number of companies have been recruited.

Ensuring the availability of active shutter glasses will obviously be one of the main concerns for AMD but most certified monitors will come with at least one set of glasses. Considering nearly all of the big names in the active shutter glasses market seem to be supported through HD3D, there should be no shortage of high quality glasses from the likes of XpanD and Bit Cauldron.

AMD’s graphics cards will also be compatible with Bit Cauldron’s excellent HeartBeat technology that allows for the virtual elimination of the sync issues that sometimes plague active shutter glasses. Bit Cauldron’s BC5000 glasses are actually some of the first to boast AMD certification.

Since there aren’t native stereo 3D drivers available from AMD, middleware partners are expected to provide third party driver wrappers which enable support through programs that piggyback off of the Catalyst drivers. DDD and iZ3D have been releasing these compatible wrappers for quite some time now and together they provide support for some 400 games. Both allow stereo 3D to be added to games and movies which don’t natively support depth perception.

This is actually a significant risk since AMD has very little control over the quality and compatibility of third party software. These driver wrappers very rarely carry WHQL certification which could lead to additional conflicts as well. From our experience, the DDD’s TriDef and iZ3D’s own software don’t play nice together so users will likely have to choose between one or the other. In addition, the number of native stereo 3D games that support AMD solution at this point is precisely zero which is why AMD needs to count on these driver wrappers from third parties. AMD simply has no plans to implement their own stereo 3D drivers.

One of the most important things to remember is that even though AMD is pimping HD3D, they are not actually the ones doing the vast majority of development. This has already led to both iZ3D and DDD releasing driver wrappers that work equally well with NVIDIA cards as they do with AMD’s products. One good example of this is the recently announced Viewsonic V3D241WM-LED which uses wired shutter glasses and needs a specific iZ3D driver wrapper to function properly with GeForce and Radeon products.

Being an open standard, HD3D can’t really be classified as AMD’s technology and believe it or not, support for Radeon graphics cards from DDD, iZ3D, XpanD and all the other companies listed above isn’t anything new. AMD is basically helping developers implement support for their graphics cards without introducing a proprietary standard. Due to a lack of content and quality control, there may be some issues with this approach but only time will tell whether or not it can be counted successful.

 

[SKYMTL]

Eyefinity Gets Refined

Eyefinity Gets Refined

One area in which AMD blazed a trail over the last year or so is in the surround gaming market. With the introduction of Eyefinity technology, gamers interested in using multi monitor setups no longer had to worry about spotty driver support or jumping through hoops to get things working properly. NVIDIA released their own Surround-branded multi monitor support as well but AMD hasn’t been standing still when it comes to updating Eyefinity.

Naturally, AMD sees fit to brag about their accomplishments in terms of accessibility for Eyefinity and in the coming months there will be yet more reasons to choose Eyefinity over NVIDIA Surround. Features such as a 5x1 portrait mode, enhanced bezel correction and more customization tools will soon be added but the most interesting addition will be on the Northern Islands cards themselves.

It isn’t quite business as usual on the backplate of the HD 6000 series since AMD has augmented the connector selection in order to better support Eyefinity. In order to make Eyefinity slightly more flexible, a pair of mini DisplayPort 1.2 connectors has been added in the place of the single large DP 1.1 connector which was seen on the HD 5000 series.

DisplayPort 1.2 brings one huge advantage to Eyefinity users: the ability to drive up to three monitors off of a single connector via a hub that will be sold separately. In addition, most of the upcoming monitors which utilize this new standard will have both DisplayPort inputs AND outputs so you can connect the primary display to the GPU and then use the output on the display to daisy chain other monitors together. This means a single HD 6800 series card can natively support up to 6 monitors.

Bandwidth shouldn’t be an issue either since the DisplayPort 1.2 standard effectively doubles the bandwidth of the current 1.1 standard to approximately 17 Gbit/s. This is enough to run up to four 1080P displays at a 60Hz refresh rate or two 2560 x 1600 displays off of a single connector and is sufficient to support 120Hz stereo 3D content to a single 120Hz display with a resolution up to 2560 x 1600. 3D is also supported through the included HDMI 1.4a connector.

 

[SKYMTL]

From ATI Stream to AMD APP Technology

From ATI Stream to AMD APP Technology

As AMD is moving to all but abolish the ATI brand, a number of technologies are being renamed and rationalized. The Stream name was once given to ATI’s GPGPU compute initiative to differentiate it from NVIDIA’s highly successful CUDA environment. Things are about to change….a bit.

In order to carry on Stream’s torch, AMD Accelerated Parallel Processing or APP has been created. Yes, the whole “App” moniker has been used so much that it’s now becoming a bit of a cliché but in this case, it seems to be aptly translated. This is now an all-encompassing term which can be used for graphics cards as well as AMD’s new upcoming generation of Fusion APUs.

AMD’s new APP SDK v2.2 is about to hit developers’ doorsteps and with it will come the integration of OpenCL computing for both x86-based CPUs and GPUs. This is one of APP’s major benefits over NVIDIA’s competing solution as it can be leveraged for a heterogeneous environment where specific tasks are sent towards the processor which will complete them most efficiently.

From our understanding, Stream’s name may have changed but its goal to deliver high performance computing on the GPU through OpenCL, DirectCompute and other APIs is still very much alive and well. If anything, it has actually expanded now that AMD is able to leverage both their CPUs and GPUs under the same programming umbrella.

 

[SKYMTL]

Meet the Cards: Reference HD 6870 1GB

Meet the Cards: Reference HD 6870 1GB

​

The reference HD 6870 looks almost like every other high end card that has been released in the last few years as it makes use of a blower-style heatsink design with a single 80mm intake fan. At 9 ¾” long, it also happens to be one of the larger sub-$250 cards currently on the market

​

​

Even though AMD has all but abolished the ATI name, the HD 6800 series of cards continue on with the usual black and red color scheme. In the HD 6870’s case, it does away with the Batmobile-like intake vents of its predecessor for a black / red sandwich look that actually goes quite well with the design intent of the main decal.

Since this card can consume up to 151W in its reference form, a pair of 6-pin PCI-E connectors are used and have been placed on the edge of the PCB. AMD has also decided to keep the single Crossfire connector from the HD 5700 series which means only two cards can be used at the same time.

​

As we discussed in the Eyefinity section, the HD 6870 uses a new output connector layout in order to facilitate Eyefinity compatibility. It includes a pair of mini DisplayPort 1.2 connectors which can be used with monitors via an adaptor that some board partners will be including with their cards. There is also a single HDMI 1.4a outputs as well as two DVI connectors.

​

At 9 ¾” long, the HD 6870 is substantially longer than the 8 ¼” GTX 460 1GB but this shouldn’t cause any compatibility issues in standard sized ATX cases.

​