Right. First, before we dive into the photo-shoot and benchmarks, we are going to walk through the reference architecture that is Cayman, the GPU embedded on the Radeon HD 6950 and 6970 graphics cards.
An Architecture Change
The previously released cards in the 6800 Barts series, the 6850 and 6870, were merely a small architectural optimization/tweak over the last generation Cypress architecture. With the Cayman GPU, things have changed a little bit as the fundamental section of the GPU, the Shader processor setup underwent a significant change, and we are still debating whether or not it was a good one.
AMD moved from a VLIW5 (also known as VEC5) towards a VLIW4 SIMD shader processor setup. We are not going to discuss the VLIW4 thread processor setup in much detail but basically what this means is that AMD went from a VLIW5 configuration, that used four simple SIMD units and one complex t-unit (transcendental unit) in order to build a stream processing unit, to a VLIW4 configuration that uses four stream units which feature equal capabilities, two of them being assigned with special functions.
AMD however claims that this change will bring them 10% more performance over the previous thread processor setup, better scheduling and register management. We think it was merely a design change to save on the number of transistors which you can re-use to add more shader processors on the processor die.
Next to this rather significant change, are some more changes to be found on the graphics card. It has upgraded render back ends (ROPS) with a redesigned Z-Stencil and ROP unit architecture consisting of 128 Z/Stencil ROPs, and 32 color ROPs, up to 2 times faster in 16-bit integer operations and two to four times faster in 32-bit floating point operations which you will have in AA performance, much faster GDDR5 memory and we also spot a series of improved compute features that will help out in the performance in that segment.
One other detail that you might find interesting is that when you look at the block diagram, you'll notice that the GPU pretty much looks like a dual-core processor. AMD calls this dual graphics engines. Anyway, have a peek at the block diagrams if at all interested.
Alright, some more generic information to grasp. The Cayman GPU itself is based on a 40nm fabrication process and harbors a blistering 2.64 Billion transistors. The graphics engine can have up-to 24 shader clusters, with each engine holding 64 shader processors. Do the reverse math and you'll quickly learn that the most high-end GPU will have a count of 1536 shader processors. A bit of an unusual number and we just wonder if there isn't more of them to be found inside that die really.
The Cayman chip has up-to 96 Texture Units and can produce 2.7 TFLOPs of single precision performance.
Memory wise AMD of course stuck to its fine working GDDR5 setup, and yes... it is still based on a 256-bit memory bus. They did increase the effective data rate though, the fastest product today will run a 5500 MHz (effective) memory clock frequency. We continuously say "effective" as GDDR5 memory is quad data rate memory. So 5500 MHz in fact is 4x 1375 MHz.
Very notable is that these cards all come with 2 GB of memory, that's both the R6950 and R6970. But let's look into the specifications a little more in-depth.
So then, based on Cayman, initially two products are thus released, the Radeon HD 6950 and 6970. Here are the particulars placed into a table.
Radeon HD 6950
Core Clock / MHz
Memory Clock / MHz
The Radeon HD 6950 comes armed with 1408 Shader processors, thus 22 SIMD based shader clusters, split up in a two-fold engine. The domain and shader clock is locked in at 800 MHz. The card comes paired with 2 GB of memory clocked at (effective) 5000 MHz. The TDP of this product is 140 W which can be extended to 200 W board power with a new feature called PowerTune which we'll explain later.
Let's have a quick comparative overview of some of the specifications representing a certain scope of other performance parts:
Radeon HD 5850
Radeon HD 6850
Radeon HD 6870
Radeon HD 6950
Radeon HD 6970
1 GB GDDR5
1 GB GDDR5
1 GB GDDR5
2 GB GDDR5
2 GB GDDR5
Both cards are of course up-to-date DX11 class products with a couple of new features.
The MSI R6970 Lightning card is clocked at 940 MHz on the core and 5500 MHz on the memory.
Features wise, both graphics cards will be very similar to the last generation products and are merely advanced, updated models. However, some features have been updated, like DisplayPort which now follows 1.2 interface specification, HD3D, UVD3 and HDMI 1.4a are introduced. We also spot a new Anti-aliasing mode (Morphological AA), better Anisotropic filtering and improved Tessellation performance, up-to twice the performance of that of the 5000 series.
MSI Radeon R9 390X Gaming 8G OC review We review the MSI Radeon R9 390X Gaming 8G OC edition. The cards equipped with that renamed Hawaii GPU, now called Grenada. Thanks to a huge triple slot air cooler based on the TwinFrozr V design the ...
MSI Radeon R9-290X Lightning review This review will cover the MSI Radeon R9-290X Lightning, this graphics card is considered the beast that rules them all. With a PCB redesigned from ground up and usage of only the best components ..
MSI Radeon R9-290X Gaming OC review We review the MSI Radeon R9-290X Gaming OC. The cards equipped with that Hawaii GPU now is proper cooled thanks to the TwinFrozr cooler and steadily running and purring at just over 70 Degrees C, that...
MSI Radeon R9-270X HAWK review In this review we take a peek at the Radeon MSI Radeon R9-270X HAWK. Armed with a customized PCB and their TwinFrozr model IV based coolers they factory overclocked the product.