Thursday, July 11, 2013
Posted by Berny Polanía | 1:23 PM | A short review for new graphic cards
The GeForce GTX 770 can best be described as a GeForce GTX 680 with a slight overclock (up 40 MHz to 1046 MHz) and a more notable boost (up 250 MHz to 1752 MHz). Introduced at a compelling $400, the 770 doesn't sport an exciting new GPU, but it's certainly a value leader compared to AMD's Radeon HD 7970 GHz Edition, which tends to cost $40 to $50 more.
At that price, the refreshed GK104-based board earns our recommendation.
In comparison, the GeForce GTX 780 is a de-tuned Titan with two of its SMXes disabled.
The GPU still wields 2304 CUDA cores, though. In the process of slimming GK110 down, Nvidia axes the chip's advanced FP64 capabilities, dropping back to 1/24 of the FP32 rate.
Gamers won't really care, though.
On the plus side, GeForce GTX 780 costs $350 less than Titan.
Even still, this $650 board is too expensive to garner any sort of value-oriented recommendation. It is an interesting option for builders looking to cram lots of performance into a small place with modest power use.
Nvidia is also getting more aggressive with the pricing on some of its existing products, particularly the GeForce GTX 670, which is down $20 to $350.
The GeForce GTX 680 dropped $30 to $420, but with GeForce GTX 770 selling for less, informed gamers won't be touching that one.
On a side note, GeForce GTX 690 and Titan are both tough to find, with single models going in and out of stock on Newegg.
As far as AMD's products go, we're seeing $10 drops on the Radeon HD 7970, 7950 Boost, 7870 2 GB, 7850 1 GB, and 7790.
Unfortunately, none of those adjustments take those models close enough to competing GeForce models.
In addition, we're surprised to see the average price of low end Radeon HD 6450, 6570, and 6670 DDR3 cards increase about $10.
Thursday, November 22, 2012
Posted by Berny Polanía | 9:25 AM | Evga Geforce Gtx 690 Signature 4096 mb Gddr5 512 bit Graphic Card
The Geforce® Gtx 690 is Nvidia’s fastest graphics card ever.
It combines the groundbreaking new Geforce architecture with raw dual gpu power to provide truly game changing performance.
Get superior gaming technology that’s meticulously designed inside and out for the fastest, smoothest, richest gaming experience.
It is one of the fastest graphic cards and gives a truly amazing performance.
This wonderful graphic card is designed in such a way that it gives the smoothest and fastest gaming experience.
It has a base clock of 915 MHz and boost clock of 1019 MHz, while memory clock is of 6008 MHz.
It is a Pci express 3.0 graphic card with 3072 cuda cores.
Key features :
Nvidia Smx engine
Nvidia Gpu boost
Nvidia adaptive vertical sync
Supports four concurrent displays; three dual-link dvi connectors, one mini display port 1.2*
Microsoft DirectX 11 with Direct Compute 5.0 support
Nvidia PhysX Technology
Nvidia 3D vision ready**
Quad Nvidia Sli ready
Nvidia Cuda technology
Pci express 3.0 support
Opengl 4.2 support
Memory detail : 4096 mb gddr5
Memory bit width : 512 bit
Memory clock : 6008 mhz
Memory speed : 0.33 ns
Memory bandwidth : 384.52 Gb/s
Core specs :
3072 Cuda Cores
Base clock : 915 MHz
Boost clock : 1019 MHz
Bus : Pci-express 3.01
Quad sli ready
Texture fill rate : Tba Gt/s
Monday, September 24, 2012
Posted by Berny Polanía | 12:22 PM | Msi launches Gtx 650 power edition graphics card
Mainboard and graphics card maker Msi has announced the release of the Gtx 650 power edition graphics card.
According to the press release, the Msi Gtx 650 boasts a number of new features, such as the Msi power edition architecture, which provides overclocking support and transthermal cooling that allows better control over the temperature of the card.
Msi also claims that the Gpu is built with Msi Military Class III components, which would give it high quality and stability.
The card also allows users to install an extra fan to it for better cooling.
Msi''s Gtx 650 power edition is based on the Msi Power Edition architecture : support for Afterburner's Triple Overvoltage function, adjustable Gpu/ memory/ Pll voltages, completely liberating the graphic cards capacity, increasing overclockability by up to 27 percent.
The new Pwm design strengthens power capability 33 percent more than the reference board maintaining stability during overclocking.
Msi Gtx 650 Power Edition features the TransThermal cooling design that allows users to tailor the fan setup to the outside environment.
Double airflow mode uses a special mechanical design for attaching the fan to the existing heat sink module and increasing airflow to effectively reduce the Gpu temperature.
The external fan can also be placed next to the existing module in the dual fan mode.
The increased coverage improves cooling for the power supply module and components.
The Gtx 650 Power Edition also comes with Msi's proprietary dust removal technology.
The fan automatically runs in the reverse direction at start-up every time to remove dust and optimise cooling performance.
Msi’s Propeller Blade technology supposedly increases airflow by 20 percent compared to conventional fans and also expands the cooling area.
Effective temperature can be reduced by up to 9 degrees Celsius and noise levels can be reduced by up to 16.8dB.
The Msi Gtx 650 is made using Msi Military Class III materials.
A third-party laboratory simulated the user environment and conducted seven rigorous tests on compliance with the Mil-Std-810g standard.
Msi Gtx 650 power edition uses Tantalum Core Hi-c Cap and SFC can generate 30 percent more power. Solid state capacitors also have a service life of up to 10 years.
The choice of materials reflects Msi's commitment to its products.
Regardless of the overclocking or the gaming environment, MSI aims to provide consumers with a smooth and dependable environment.
The Gpu also has the latest drivers available on Nvidia’s website, which offers more optimisations for current-generation games.
The driver update comes highly recommended for the upcoming games Borderlands 2, and the expansion pack to World of Warcraft.
Monday, September 17, 2012
The bulk of products that Advanced Micro Devices will be shipping in 2.013 will be made using 28nm process technology, according to plans for the year 2.013 that were disclosed at the company's Financial Analyst Day.
AMD expects to roll-out three brand-new architectures next year : steamroller high-performance x86, Jaguar low-power x86 and Sea Islands family of graphics chips.
All product lineups that Amd promises to release next year will be substantially different from what is available today, hence, they will not be just product shrinks or refreshes.
28nm process technology alone is likely to allow Amd to boost performance and feature-set of its chips without increasing power consumption compared to existing solutions that are made using 32nm or 40nm fabrication processes.
Unfortunately, Amd did not disclose which of its 28nm chips will be made at Globalfoundries and which at Taiwan Semiconductor Manufacturing Company.
The new Steamroller x86 high-performance cores will improve performance by boosting efficiency of execution and will also include design elements from low-power architectures to reduce power consumption.
While it is unknown whether code-named Kaveri accelerated processing unit will turn the page for Amd in terms of x86 performance, the company hopes that further integration of x86 and stream processing architectures will boost overall user experience.
The forthcoming Jaguar low-power x86 architecture is expected to boost performance while trimming energy consumption, which will let Amd to better compete for tablets and similar devices.
Thanks to thin process technology, Amd's code-named Kabini and Temash APUs will also integrate input/output capabilities, which will greatly simplify system design.
Finally, the code-named Sea Islands graphics processors promise to bring "major GPU architecture enhancements" and skyrocket performance for both graphics and computing.
What is a bit surprising is that a high-performance FX-class x86 chip featuring Steamroller cores is absent from the 2.013 plan, but the Steamroller will still become available with the code-named Kaveri accelerated processing unit.
Apparently, Amd plans to continue selling its enthusiast-class FX chips code-named Vishera with Piledriver cores due to be introduced in 2.012 next year as well.
Tuesday, August 21, 2012
Amd has released a card that's just, well, better: the Amd Radeon Hd 7970 GHz Edition.
Since it released the original Radeon Hd 7970, in december last, the top single-GPU graphics card crown has slipped from its once mighty, and once mightily expensive, head.
Nvidia waited, holding its breath until March, when it released its top consumer spin of the new Kepler architecture, the Nvidia GeForce Gtx 680.
That came in slightly undercutting the Radeon card and just about pipping it to the benchmark crown in most of our tests.
Not just that but the Nvidia GTX 680 managed it with a far smaller and more efficient GPU.
To most of us that came as something of a surprise.
AMD traditionally had the reputation of releasing the more efficient GPUs of a generation with Nvidia usually the ones employing a brute force approach with a big, hot and powerful chip at the heart of its latest architecture.
That switch wasn't the only one of this latest generation either.
AMD's Graphics Core Next (GCN) architecture which makes up the Southern Island series of HD 7000 graphics cards was conceived with a view to creating a much more compute focused card.
Previous to this architecture Amd had focused on creating pure gaming cards with only a faint nod towards GPGPU processing.
Nvidia, on the other hand, chose performance per Watt for its consumer cards and sacrificed the GPGPU computing power that had been central to the Fermi architecture.
This cut-down GPU though was still a great gaming chip, as the Nvidia GeForce GTX 680 showed, leaving the Tesla team, with the GK110 K20, to the pro-graphics compute shizzle.
Still, it managed to best AMD's top card in most gaming benchmarks leaving it with the runners up prize.
The AMD Radeon HD 7970 GHz Edition is the Texan company's counter offer to gamers in light of the slight lead Nvidia managed with the GTX 680.
And to be fair, it's come up with quite an interesting proposal.
The salient point here though is that AMD has reclaimed the outright single-GPU performance crown of this generation.
Obviously that's discounting the twin-GPU toting Nvidia GTX 690, which has almost cheated its way to the top of the graphics card pile.
Though AMD had planned an assault on even that position, promising us the card once code-named as New Zealand.
You know, the place that has two Southern Islands making up the country.
Still, for whatever reason that card is off the table now with AMD focusing its efforts on bolstering the Tahiti GPU-sporting HD 7970 and making it the top single-GPU card around.
How has it managed this feature?
Well, the simple answer is overclocking.
We told you it was simple.
But this isn't quite the traditional card speed bump, or manufacturer-supported factory overclocked card you might normally expect.
Nothing has actually changed regarding the GPU itself.
AMD is still taking them from the same wafers, with the same designs, but it has taken the six months of learning that went into the subsequent Pitcairn and Cape Verde GPUs and applied it to the flagship chip.
Basically the GPU manufacturing process has matured and so has the technology and, most importantly, the software.
This has enabled AMD to get more top-end Tahiti cores out of the wafers and so is offering the AMD Radeon HD 7970 GHz Edition with a clockspeed of, you guessed it, 1GHz.
Though that's not strictly true.
In reality the clockspeed you're realistically going to get when gaming is actually 1.05GHz.
That extra 50MHz coming from something AMD is calling PowerTune with Boost.
The idea being similar in concept to Nvidia's GPU Boost, with some new hardware on the PCB monitoring the voltage and offering extra power, up to that 1.05GHz mark, when the thermal headroom allows.
In our testing that was all of the time, even with the standard reference cooler design.
Our reference AMD Radeon HD 7970 GHz Edition sat at a stable 1.05GHz for the entire time the GPU was pushed to 100% load.
The only other real addition of note to the card is the 3GB of GDDR5 graphics memory has been given a bit of a speed boost too.
The stock card has a 1,375MHz speed for its memory while the AMD Radeon HD 7970 GHz Edition comes with the VRAM running at a healthy 1,500MHz.
That's not all though as the type of memory used has been changed too.
This time it is using the faster 6Gbps VRAM over the 5.5Gbps chips it had been rocking, the same speedy memory the Nvidia GTX 680 is sporting then.
Thanks to AMD's faster 384 bit memory bus though the Radeon HD 7970 GHz Edition has a massive 288GB/s memory bandwidth compared with the 192GB/s of the top single-GPU Nvidia card.
That means at the higher end of the resolution spectrum the AMD card now seriously has the edge.
This video card is great because it has :
Lightning fast performance
Great compute performance
Top GPU on the block
Ant it is incredibly expensive
And has a great value with it
GPU - Tahiti XT
Process - 28nm
Clockspeed - 1,050MHz
Memory capacity - 3GB GDDR5
Memory bus - 384-bit
Thursday, August 2, 2012
Posted by Berny Polanía | 5:35 PM | Nvidia Dual GeForce Gtx 580 - Asus Mars II
Asus has never been a company to shy away from offering over-the-top, custom products that command a price premium.
Here's proof. Perhaps it’s because the company is large enough, and thus, has the resources to cater to virtually all market segments, big or small.
Or perhaps the top brass at Asus just likes inducing slack-jawed stares from PC enthusiasts. Whatever the reason, we hope they keep it up because testing this kind of stuff and showing it to you all is freakin’ fun.
A case-in-point is the Asus MARS II graphics card, which packs a full GeForce GTX 580, dual-GPU SLI configuration onto a single PCB.
A quick glance at this monolith should reveal to even the most casual PC users that the MARS II is no ordinary graphics card.
Heck, there isn’t a single thing about the Asus MARS II, from its specs, to its sheer size, and its huge price tag, that can be considered mainstream.
The MARS II is as over-the-top and as custom as they get and it’s also going to be a limited edition, with only 999 cards slated for production.
Luckily, one of those 999 cards (number 545 to be exact) found its way into the HH labs and we’re able to give you the full scoop on this beast.
Dissect the Asus MARS II and evaluate its performance, here’s a look at its packaging and accessory bundle.
For the most part, it’s pretty standard fair.
Included with the card are a couple of 8-pin PCI Express power adapters, a DVI-to-VGA adapter (for the insane user who uses a $1500 graphics card with an analog display), a driver and utility CD (with Asus' excellent GPU Tweak tool), a RoG case badge, a speed-setup guide, and an extended length SLI bridge.
In addition to the aforementioned items though, Asus also throws in a couple of noteworthy goodies.
First is an aluminum plate that’s individually laser-engraved and sequentially numbered to certify the card’s limited edition status.
Second, there are a couple of spacers that Asus calls “PCIe Sustainers”.
These spacers adhere to the bottom of the card and rest upon adjacent expansion slots to ensure the MARS II sits perfectly straight and upright in its slot.
Why would you need spacers to help keep the card sitting upright?
Because the thing weighs over 5lbs.
Model MARS II/2DIS/3GD5
Graphics Engine NVIDIA GeForce GTX 580 x2
Bus Standard PCI Express 2.0
Video Memory 3GB GDDR5
Engine Clock 782MHz
Memory Clock 4008MHz (1002MHz GDDR5)
Memory Interface 384-bit x2
D-sub Max. Resolution 2048 x 1536
DVI Max. Resolution 2560 x 1600
D-sub Output Yes x 1 (via DVI-to-D-sub adaptor)
DVI Output Yes x 2 (Native dual link DVI x2)
HDCP compliant Yes
HDMI Output Yes x 1 (Native HDMI 1.4)
Display Port Yes x 1 (Regular DisplayPort)
Adaptor/Cable Bundled 1x Limited Edition aluminum ID plate
2x 8-pin to 6-pin Power Cable
1x Exclusive 12cm SLI bridge
Software Bundled ASUS utilities and drivers; GPU Tweak
Dimension 13 inch x 6.2 inch x2.5 inch(3slot)
To have the best cooling performance ASUS ROG MARS II extends the Fan sink to 3 slot, please check your motherboard slot space and Chassis Dimension before Purchasing.
ASUS ROG Rampage III Extreme or Rampage III Extreme Black Edition Mainboard is recommended for best Synergy.
Sunday, July 8, 2012
Posted by Berny Polanía | 12:50 PM | The midrange and the low end of video cards in 2.012
If your frame rate needs are a lot more modest (or your budget is under $200), the two cards Amd announced in february are worth considering, as well.
The Radeon Hd 7770 Ghz edition is our pick here for serious gamers.
At $159, it's notably pricier than the $109 Radeon Hd 7750.
But the HD 7770 offers much more gaming muscle, making it a good choice for those who game at resolutions of 1,680x1,050 or lower, and who don’t want to dial back many in-game settings.
Gamers who intend to play at 1080p (true high-def) resolutions on a single monitor and who are willing to switch off some in game eye candy to get smooth frame rates at a fair price will find a lot to like in the HD 7770 card, as well.
Amd, though, calls the “sweet spot” of 2.012 gaming performance its cards that lie between the ultra enthusiast 7900 series (priced above $400) and the under-$200 7700-series mainstream cards.
And here, we have to give Amd the nod.
We think the company's right on that score, with its pair of 7800-series cards announced the first week of March 2.012.
The Radeon HD 7870 GHz Edition, at $349, easily bests Nvidia’s comparably priced Geforce GTX 500-series offerings, and in some cases it bests the $500 Geforce GTX 580.
At $249, though, the somewhat lesser Radeon Hd 7850 is our favorite pick of all of Amd’s current release wave.
It has the same features as the other 7000 series cards; it only requires a single six-pin power connector; and it sips much less power than comparably performing Nvidia cards.
Oh, and did we mention that its performance in our tests was mostly level with (and sometimes better than) the much costlier Geforce Gtx 570?
That was the clincher with this Editors' Choice-winning card.
If you’re looking for the real sweet spot of value and performance in a graphics card in the spring of 2.012, the Radeon Hd 7850 is it, in our opinion at least until Nvidia unveils its own new cards in this price range.
More on that, we suspect, later in the spring.
Saturday, June 30, 2012
Posted by Berny Polanía | 12:30 PM | Sapphire Amd Radeon Hd 6850 1GB Pci-E Video Card
The Sapphire Hd 6800 series features the second generation of graphics technology from Amd designed to support all the features of Microsoft DirectX 11 and DirectCompute 11.
The new Gpu’s are built in Amd’s 40nm process to deliver high performance with low power consumption.
Advanced hardware features include an enhanced tesselation unit, dedicated hdr texture compression, multi-threaded communication with system Cpu cores, embedded support for many of the DirectCompute 11 instructions, 3D stereoscopic support and HDMI 1.4a.
Equipped with the most popular Dual Link DVI (Digital Visual Interface), this card is able to display ultra high resolutions of up to 2560 x 1600 at 60Hz.
Hdmi (with 3D)
Support for Deep Color, 7.1 High Bitrate Audio, and 3D Stereoscopic, ensuring the highest quality Blu-ray and video experience possible from your PC.
Enjoy the benefits of the latest generation display interface, DisplayPort.
With the ultra high HD resolution, the graphics card ensures that you are able to support the latest generation of LCD monitors.
Supports Digital monitors up to 1920x1200 resolution.
Amd Eyefinity technology maximises your field of view across up to three displays, fully engaging your peripheral vision.
For gamers this puts you right IN the game and for other applications it helps to increase productivity by maximising your visual workspace so that you can see more windows simultaneously.
Amd Eyefinity technology is also ideal for multimedia applications as it allows you to keep several ‘palettes’ or ‘panels’ open over a larger workspace whist you edit your images or videos.
Output 1 x Dual-Link DVI
1 x HDMI (with 3D)
1 x DisplayPort
1 x Single-Link DVI-D
Gpu : 775 MHz Core Clock
40 nm Chip
960 x Stream Processors
Memory : 1024 MB Size
256 -bit GDDR5
4000 MHz Effective
Software : Driver CD
SAPPHIRE TriXX Utility
Accessory : CrossFire™ Bridge Interconnect Cable
DVI to VGA Adapter
6 PIN to 4 PIN Power Cable
HDMI 1.4a high speed 1.8 meter cable(Full Retail SKU only)
Advanced Gddr5 Memory Technology
Gddr5 memory provides twice the bandwidth per pin of Gddr3 memory, delivering more speed and higher bandwidth.
40nm Process Technology
Cutting edge manufacturing process allows for efficient, high performance parts which use less power than their predecessors.
Amd Avivo HD (DVD/HD-DVD/Blu-ray Playback)
Watch the hottest Blu-ray movies or other HD content at full 1080p display resolution and beyond and enjoy a more feature rich experience with enhanced Unified Video Decoder(UVD)
Amd CrossFireX™ Multi-GPU Technology
Triple or Quadruple gaming power with AMD CrossFireX™ Multi-GPU Technology.AMD CrossFireX technology supports up to four graphics cards to be used in a single system.
The latest PCI-Express 2.0 specification doubles data throughput from 2.5Gbps to 5Gbps allowing you to exploit the full performance potential of your graphics card.
Amd Hd3d Technology
Enjoy the most immersive experience possible with full support for High Definition Stereoscopic 3D, a technique that presents 2D images (movies, games, photos) in a format that creates the illusion of three-dimensional depth when using compatible 3D displays / glasses / software.
Microsoft DirectX® 11
Get intense gaming performance and unrivalled image quality with stunning 3D visual effects, realistic lighting and lifelike imagery with SAPPHIRE graphics cards supporting DirectX®11.
New features in DirectX®11 such as Hardware Tesselation, HDR Texture Compression, Multi-threading support and DirectCompute bring lifelike gaming and an experience never possible on a PC before.
Supports the latest OpenGL API (Application Programming Interface) to accelerate performance of applications outside of DirectX®, i.e. Adobe Autodesk 2D/3D graphic design.
Shader Model 5.0
Supports the latest Shader Model 5.0 feature set including VTF (Vertex Texture Fetch) which is a key feature used extensively in many of the games that ship today.
Without support for this feature the game will fall back to a lesser shader path resulting in the loss of cool, lifelike effects made possible by Shader Model 5.0
Amd Catalyst Control Center™
The Amd Catalyst Control Center™ software application gives you complete control over the performance and visual quality of your SAPPHIRE AMD Radeon™ based graphics card allowing for the best experience on your PC.
Accelerated Video Transcoding
Massively parallel architecture and support for common shader transcoding instructions accelerates video encoding / transcoding way beyond what was possible with the CPU only.
(Requires application support for AMD Stream technology and Digital rights management restrictions may apply)
Dolby® TrueHD and DTSHD Master Audio™ Support
Support for copy protected, high bandwidth, 7.1 channel surround sound over HDMI.
Get a fully immersive, high definition audio experience with the latest Blu-ray movies.
Requires a DVI to Hdmi dongle on some models, an HD capable monitor and a Blu-ray drive.
Supports advanced AA (Anti-aliasing) features to maximise your enjoyment of high definition gaming.
Multi-sample anti-aliasing (2, 4 or 8 samples per pixel)
Up to 24x Custom Filter Anti-Aliasing (CFAA) for superior quality
Adaptive super-sampling and multi-sampling
Super AA (AMD CrossFireX™configurations only)
All anti-aliasing features compatible with HDR rendering
Physics processing support
Accelerates the performance of Fluid Dynamics, Cloth Simulation, Soft or Rigid Body Dynamics, Universal Collision Detection, Hair Simulation and Finite Element Analysis for the most lifelike gaming experience.
Dynamically manage power efficiency and take control of your performance with AMD’s revolutionary AMD PowerPlay TM power management technology delivering high performance features when needed and conserving power when the demand on the graphics processor is low.
This seemless technology dramatically reduces the overall power consumption allowing for break-through energy efficiency.
3rd Generation TeraScale Engine
Massive arrays of on-chip compute units offer the latest in graphics functionality and performance allowing you to enjoy what you are watching at the highest quality possible.
Amd Advanced Parallel Processing technology
Accelerate the most demanding applications by offloading parallel data compute tasks from the CPU to GPU, dividing the tasks as evenly as possible for improved system performance and reliability.
Offers an incredible choice of the latest display connectivity options natively on the board, including dual-link DVI-I,
DisplayPort and HDMIalong with adapters to ensure compatibility with your legacy analogue vga monitor.
TriXX helps users to get the optimum performance from compatible Sapphire graphics cards in any application by allowing them to monitor, tune and save different settings of key parameters for the card.
Enhanced Unified Video Decoder 3 (UVD 3)
Watch the hottest Blu-ray movies or other HD content at full 1080p (and beyond !) display resolution.
Enjoy a feature rich video experience with enhanced Unified Video Decoder 3 (UVD 3), offloading the bulk of the video decoding tasks from the CPU to the GPU allowing for a cooler, quieter and more balanced system.
Enhanced Internet Browser Applications
Not just for video and games, the power of your GPU can now accelerate the latest versions of some Internet browsers (IE9/Mozzilla FireFox) utilising Direct 2D rendering for a faster and more responsive browsing experience including high quality image rendering and media playback.
Windows® 7 support
Whql (Windows Hardware Quality Labs) Certified drivers and logos are available for all Sapphire HD2000, HD3000, HD4000 and HD5000 series cards ensuring compatibility and reliability with Microsoft Windows 7.
Amd App Acceleration
Amd App Acceleration is a set of technologies designed to improve video quality and enhance application performance.
Full enablement of some features requires support for OpenCL, DirectCompute or DirectX Video Acceleration (DXVA).
Thursday, June 28, 2012
US price (as reviewed) $369 (ex tax) MSRP
The previous Radeon HD 6800-series got our hopes up in two ways,firstly it sounded like it’d be more than a mid-range GPU that went toe-to-toe with Nvidia’s GeForce GTX 460 cards, and secondly it wasn’t all that new.
But while the Barts GPU of the HD 6800 looked more like an overclocked HD 5830 with a tweaked front-end unit, the Cayman GPU of the Radeon HD 6900 is a completely new design throughout.
Before we plunge into the details of the new GPU, let’s clear up the naming.
We’re sticking with calling these cards ATI Radeons because that’s what most people know them as, and AMD has said that it’s fine with it partners gradually transitioning from ATI to AMD Graphics until 2.011.
We still plan to switch when AMD’s Fusion APU is launched in early 2.011, as it’ll be silly to talk about a single piece of silicon that has both AMD and ATI technology.
Previously ATI has reserved the HD x900 family branding for its dual-GPU cards to give a bolder indication to the customer that a dual-GPU card should be much faster than a single-GPU HD x800-series card. However, because ATI needed the HD 5700-series to exist with the HD 6000-series, ‘Radeon HD 6800’ was already taken.
The Radeon HD 6950 2GB and Radeon HD 6970 2GB are therefore both single-GPU cards.
There is a plan to release a dual-GPU card based on two Cayman GPUs which we assume will be called the Radeon HD 6990 4GB.
With the names out of the way, let’s stuck into what’s inside the silicon that makes a Cayman worthy of its HD 6900 name.
A Dual Front-End Design
While Nvidia went nuts with its Fermi design, breaking apart the elements of a typical GPU front-end and scattering them throughout the chip,
ATI has been much more reserved.
However, we’ve been expecting ATI to get a bit more radical for a while – after all, a GPU with only one tessellator and one setup engine is starting to look a bit anachronistic these days.
While the Barts GPU added merely had a tessellator upgrade (to what ATI is called its 8th Gen Tessellator, which allows off-chip buffering) the Cayman design of the Radeon HD 6900-series has two entire front-end units.
There are some obvious advantages of having two front-end engines: you get two setup engines, two 8th Gen Tessellators, two geometry engines and the ability to send twice as much work per clock to the stream processors than before.
ATI claims that the HD 6970 2GB has up to three times the tessellation rate of the HD 5870 1GB.
These two front-end units can also load-balance the work flowing to them, and ATI has implemented ‘asynchronous dispatch’.
The new Cayman GPU of the Radeon HD 6970 2GB has two Fron-End Engines.
Asynchronous dispatch is like the ability of Nvidia’s Fermi GPUs to work on two distinct kernels simultaneously.
However, ATI says that its technology is ‘completely new in the marketplace’ as it allows multiple, different programs to execute on the GPU at the same time.
‘It’s not like other solutions where you [only] have one program that can spawn multiple kernels to run on the graphics card, you can genuinely have multiple, different applications running on the GPU at the same time’ Dave Bauman, Senior Product Manager, told us.
This should make the GPU more flexible for general-purpose work, the GPU will act more like a modern CPU.
The two bidirectional DMA (Direct Memory Access) engines of the GPU are also pitched as enhancing the GPU Compute capabilities of the GPU, as they allow two simultaneous reads or writes or a simultaneous read and write per unit.
However, we’re unsure of the need for such advanced capabilities when it comes to gaming.
A game runs as a single application, meaning that Nvidia’s technology is perfectly adequate, a game can invoke a DirectCompute kernel and throw it, as well as DirectX shader code, at Fermi GPU without concern.
Where the ATI technology will be useful is if many different applications try to use your graphics card, especially if they don’t have the courtesy to wait until you’ve finished gaming before doing so.
It’s not impossible that anti-virus applications could be written in OpenCL, for example, as virus scanning.
A short review for new graphic cards (1) Amd Radeon Hd 6990 (1) Amd Radeon Hd 7850 2 Gb (1) Asus Gtx 680-Dc2t-2Gd5 (1) Ati Radeon Hd 6970 2Gb (1) Ati Radeon Hd 7970 Ghz edition (1) Evga Geforce Gtx 690 Signature 4096 mb Gddr5 512 bit Graphic Card (1) Msi launches Gtx 650 power edition graphics card (1) Nvidia Dual GeForce Gtx 580 - Asus Mars II (1) Nvidia Geforce Gtx 590 Is the world's fastest graphics card (1) Nvidia versus Amd technology (1) Nvidia's Kepler will arrive soon (1) Sapphire Amd Radeon Hd 6850 1GB Pci-E Video Card (1) The Best Video Card Overclocking App for Windows (1) The midrange and the low end of video cards in 2.012 (1) Video card buying tips (1) What is coming in 2.013 (1)