The GF110 chip shares a lot of its roots with the GF100 core. The GPU still packs an impressive 3.2 billion transistors in a die seize 520mm squared. Its default internal configuration has been ported from GF100 in that the GPU features 512 CUDA cores or shader processors as we are more accustomed to. There are organised in a 4x16x32 block where 16 denotes the number of stream multi-processors (SM) and 32 being the number of cores on each SM. Each SM has its own L1 Cache and a common L2 Cache is shared by all 16. The GPU is still manufactured on the 40nm process.


The GTX570 has 15 out of the total 16 SM enabled for a final 480 CUDA core count. Similarly to the GTX480, it also shares the same number of texture units, which in this case is 60. These represent 32 fewer cores than the fully unlocked GTX580 and 4 texture units less. The size of the L2 Cache is reminiscent of the GTX470 coming at 640KB rather than 768KB as found on the flagship GTX580. Directly related to Cache is the number of ROPs which has been reduced from 48 to 40 on the GTX570. On the other hand, the GTX570 has been presented with a significant clock boost over both the GTX470 and the GTX480 in core and memory clock speeds to avoid “crippled” performance. Consequently, the ROPs are also clocked higher which should minimise the gap between the GTX480 and itself. Thanks to the GF110 GPU, the GTX570 also benefits from the GF104’s superior texture filtering capabilities and further enhancements have been made to the tessellation engine for improved performance in an area where Nvidia already reigns supreme.

The memory subsystem on the GTX570 is more akin to the GTX470 with its 5 memory controllers each sporting a 256MB GDDR5 framebuffer to give 1280MB of available memory. The memory interface is composed of ten 32-bit channels for a 320-bit interface.

GTX580: 16 SMs x 32 = 512 Core, 16 SMs x 4 = 64 texture units
32-bit channel x 12 = 384-bit interface, 256MB x 6 memory controllers = 1536MB memory

GTX570: 15 SMs x 32 = 480 Core, 15 SMs x 4 = 60 texture units
32-bit channel x 10 = 320-bit interface, 256MB x 5 memory controllers = 1280MB memory



The GTX570 looks to be very competitive against the GTX480, with 8.7 billion GPixels filled and 6.8 billion faster than the stock. Interestingly, the MSI GTX570 Power Edition OC has a TDP of just 180W so theoretically it will consume much less power than the cards nearest competitor on our tests today by some 70W and 39W less than a standard GTX570 while maintaining a higher pixel fill rate than the reference card. Looking at the specification, the only problem I can foresee is the 1.25GB of GDDR5 which may cause some issues at the highest resolutions.



Let's take a look at the cards feature set...