AMD Radeon HD 7850 and 7870 review

Posted by Hilbert Hagedoorn on: 03/04/2012 02:00 PM [ 0 comment(s) ]

Graphics engine architecture

The Graphics Engine Architecture

With the initial release of the 7000 series AMD moved away from the VLIW5 and VLIW4 architecture we have seen in the last generation of products. If anything, VLIW4 has shown certain inefficiencies in the Radeon HD 6900 series and while VLIW designs are fine for graphics, they are not so grand for computing.

Both the Radeon HD 7700 and 7800 series are also based on the very same GCN architecture the 7900 series is based off.

GCN is short for Graphics Core Next architecture and the architecture building block has changed significantly to remove certain inefficiencies seen in the VLIW architecture. A GCN in essence is the basis of a GPU that performs well at both graphical and computing tasks. For the compute side of things the new GCN Compute unit model has been introduced, it is designed for better utilization, high throughput and multi tasking, e.g. performance, performance, performance.

So your basic new Shader cluster is one called a (GCN) Compute Unit:

Non-VLIW Design
16 wide SIMD units
64 KB registers / SIMD unit

Now if we take 4 of these SIMD units that will be the basis of one Compute Unit (CU), each SIMD unit is 16 wide, times four per compute unit means that each CU has 64 shader processors.

The Pitcairn PRO GPU has 16 Compute Units meaning 64SIMDs x 16 CUs = 1024 Shader processors (for the R7850).
The Pitcairn XT GPU has 20 Compute Units meaning 64SIMDs x 20 CUs = 1280 Shader processors (for the R7870).

Engine has Dual Geometry engines / Asynchronous Compute engines
32 color ROPs per clock cycle
Engine ties to 512KB R/W L2 cache
Pitcairn GPU has up-to 20 Compute Units

The Graphics Core Next Compute Unit (CU) has about the same floating point power per clock as the previous one (i.e. Cayman). It also has the same amount of register space (for the vector units). Each CU also has its own registers and local data share.

Radeon HD 7800

GCN is more efficient since it does not require instruction level parallelism (we assume it costs some more area/transistors as well). The outcome, compiling also becomes much more uncomplicated and that means more efficiency and thus there it is again, better performance. GCN is all about creating a GPU good for both graphics and computing purposes.

24 pages « 2 3 4 5 next »

AMD Radeon HD 7850 and 7870 review
We review the AMD Radeon HD 7850 and 7870. These two new mid-range cards are going to shift the dynamics in the graphics arena alright, as the entire package including performance is really impressive for the 7800 series. A product series that is to replace the 6800-series performance-wise, it is based on AMD's 28nm process and of course the latest Graphics Core Next GPU architecture.

AMD Radeon HD 7750 and 7770 review
It's now February 2012 and AMD thinks they have a new '5770' in their hands. The codename is 'Cape Verde' for the GPU, and the graphics cards deriving from them are the Radeon HD 7750 and 7770 One GHz edition. This is not a refresh it is a completely new GPU based on the same technology that powers the R7900 series, the GCN architecture. Head on over to the next page where we'll meet and greet Cape Verde, aka Radeon HD series 7700.

AMD Radeon HD 7970 review
We review the Radeon HD 7970. Injected in the 499 EUR / 549 USD price tag bracket the product will have to compete directly with the equally expensive GeForce GTX 580, it will actually be a decent notch better then that IMHO. The results that you'll witness today will not dishearten. Where it matters (the latest and newer games) the Radeon HD 7970 will be a good 20%, 30% sometimes even 40% faster then the competition, and in the world of enthusiast graphics performance that's what we call, a product with a little extra booty.