NVIDIA Officially Launches The GeForce GTX 780 Graphics Card With GK110 Core Architecutre
NVIDIA, today launches their latest and greatest GeForce GTX 780 graphics card which is built using the same DNA “core” used on the GeForce GTX Titan and Tesla K20X graphics accelerators. The GeForce GTX 780 comes with the GK110 architecture that provides supercomputing performance to PCs.
Titan’s Little Brother Has Arrived – Meet The GeForce GTX 780
On the technical side, NVIDIA made a few changes with the GK110 core architecture. It still features 7.1 Billion transistors and comes with 4.5 TFlops of computing performance. The difference in the core is the GK110 version used on the GeForce GTX 780 “GK110-300″ comes with 2304 cores which means that out of 15 total SMX units only 12 were kept operational. The GeForce GTX Titan has 13 SMX units which amounts to 2668 Cores while a fully enabled GK110 GPU features 2880 cores, NVIDIA hasn’t released a GPU which comes with the fully enabled version.
Coming back to the specifications, the 2304 cores are coupled by 192 texture mapping units and 48 ROPs (Raster operators). The GeForce GTX 780 has 3 GB GDDR5 memory compared to the 6 GB GDDR5 memory on GeForce GTX Titan.
Memory operates along a 384-bit wide interface while clock frequency is
configured at the reference 1502 MHz or 6 GHz effective. The core clock
is maintained at 863 MHz while GPU Boost 2.0 is clocked at 902 MH z. The GeForce GTX 780 pumps out a bandwidth of 288.4 GB/s which is the same as the GeForce GTX Titan but a boost over the 192 GB/s bandwidth of the GeForce GTX 680.