It’s not the first time that NVidia is doing a “One more thing”
after releasing the mainstream of graphics cards. Initially the
20×0 graphics cards family has been released earlier this year
and probably many thought: Is this all you got?

Well no, as you can see, NVidia has once again saved it’s big bang for the holidays. The Titan RTX is a real beast when it comes to pure calculation power. Featuring 576 multi-precision Turing Tensor Cores, providing up to 130 teraflops of deep learning performance you can just imagine how much calculation power that beast has. And get this: You can buckle up two of these cards to double the fun using NVLink. But as we learned, the RTX family offers ray-tracing capabilities where the Titan RTX is using 72 Turing RT Cores, delivering up to 11 GigaRays per second of real-time ray-tracing performance. Even these are incredible numbers compared to the mainstream cards. While the mainstream cards come up with 12GB of graphics memory, a Titan RTX is housing 24GB of high-speed GDDR6 memory with 672GB/s of bandwidth which is two times the memory of previous-generation TITAN GPUs to fit larger models, textures and datasets. So even the most memory-hungry games graphic-wise won’t have any trouble moving in their stuff. The NVLink provides up to 100GB/s link speed between the two cards which should suffice to provide even the most complex distribution of memorybased and computing-based operations.

So although this card has nice specs, the price isn’t as nice as expected. Compared to a 15% gain over a 2080 RTX Ti the preice has doubled, demanding 2500 US$ from you to call one of these copper-colored beasts your own. Going SLI (NVLink) you would have to fork over a whopping 5000 US$ providing you with 48GB RAM and a 260 TFLOPs calculation power.

For gaming this is truly a bit oversized but on the other hand there are enough people willing to spend thousands of dollars just to have the fastest computer to date.