The HD7900 series will feature the new Tahiti core which is the successor to the Caymen core of the HD6900 family. The Caymen core was based on a 40nm manufacturing process so it is fair to say that perhaps the biggest improvement to the Tahiti core is the reduction to 28nm process. This has enabled AMD to cram 4.31 billion (4,312,711,873) transistors into the new Tahiti core as opposed to the 2.64 billion on the HD6970 Caymen XT.
Comprising of what AMD calls 'Graphics Core Next (GCN)' archetecture which is optimised for heterogeneous computing, the new Tahiti core has upto 32 CU's (compute units) with dual geometry engines along with 8 back-ends comprising of 32 colour ROPs per clock and 128 Z/Stencil ROPs per clock. Each CU has a Vector unit and Scaler Co-Processor and can execute instructions from multiple kernels.
The VLIW4 SIMD has evolved to the GCN quad SIMD which equates to similar computational power yet grants better performance, threading and simplification. The GCN also has a new Cache hierarchy with 16KB instruction Cache and32KB Scaler Cache shared between 4 CU's. The L2 Cache is accessible by each CU with it's own separate process to ensure coherency. These changes increase the GFLOPs by 140%.
Tesselation on the Tahiti core is also enhanced by increased Vertex re-use off-core which theoretically at least, means the 7900 series has 4 times the throughput of the 69XX series. The are improved ansioscopic filtering algorithms too which come at no performance cost allegedly. These new algorithm improvement will allow the use of Forward Rendering with 1000's of potential light sources and will also mean MSAA can be rendered easily.
Tahiti XT Schematic
Looking at the schematics above, we can identify the bulk of the GPU is taken up with the 32 CU's. Above these are the geometry engines. To the left and right are the 8 Render back-ends, 4 each side. There's 768KB of L2 cache along with 6x64bit dual channel memory controllers making a 384-bit GDDR5 memory interface capable of delivering up to 264GB/sec. All of technology coupled together is transferred through a fully compatible PCI Express 3.0 x16 bus interface. Have no fear though because the cards are backwards compatible with PCIe 2.0. Note that 256 Stream Processors are disabled for the Tahiti Pro Core. Whether this is at the hardware or simply a software level remains to be seen. If it is software then a BIOS flash may see the Pro's becomming XT's again?
Comparing the HD7950 against it's bigger brother, the HD7970, the 7950 has 256 less Stream Processors with 16 less Texture units. Obviously this equates to less Pixel and Texture fillrate and coupled with the slightly lower clocked memory of 5000MHz (effective) as opposed to the 5500 of the HD7970, performance is due to take a hit. The power requirements of the HD7950 are however slightly lower in that it has a claimed TDP of 200w compared to the 20% higher TDP of the HD7970.
Comparing the HD7950 to the green team it is interesting to note that the HD7950 has more the double the texture fillrate of the GTX570 and a 40Billion/sec advantage over the GTX580. The NVIDIA cards do however have the upper hand in pixel fillrate but by a much smaller margin. AMD card have traditionally been clocked higher than their NVIDIA counterparts and we see no change in that here with the lowest clocked HD7950 (800MHz) still surpassing the 772MHz of the GTX580. The memory clockspeed is also higher and now with a matching 320-bit memory interface, the AMD cards are finally on a level playing field. this equates to a much higher memory bandwidth than NVIDIA.
Let's take a look at the cards features...