If you've never heard of TITAN before, then allow me to freshen up your knowledge on supercomputers. TITAN is the world's fastest supercomputer located in Oak Ridge National Laboratory and it's powered by nVidia's Tesla® GPU Accelerators and nVidia's Kepler™ architecture.

Now, nVidia harnesses this technology and power and squeezes it inside their new consumer GPU accelerator, named GeForce GTX TITAN

The new accelerator is essentially a consumer version of their Tesla K20X board with the same 2,699 CUDA® core count and sporting 4.5 teraflops of horsepower which will lead gamers around the globe to salivate just in the prospect of getting their hands on one of those beasts (let alone three of them in SLI). The Titan comes 6GB of 364-bit GDDR5 memory, clocked at 6GHz with a total bandwidth of 288.4GB/s.

As there are no benchmarks out yet, we cannot be sure about the exact hands-on numbers and fps counts, but with these kinds of specs, it's safe to assume that nVidia will reclaim the speed king crown with this board. On the "downside" the GTX TITAN will only find home to those who can reach deep in their pockets to dish out the $1300 that its price tag demands. On top of that, the Titan's power requirements reach a whopping 250 Watt count. Running three of those together consume power equivalant to a lush chandelier.

Apart from its stellar numbers in gaming, the TITAN will also peak the interest of Cuda and OpenCL programmers who now can posses unprecedented processing power on their fingertips.