What’s The Deal With Nvidia’s Ridiculous $4000 Titan RT Graphics Card?

What’s The Deal With Nvidia’s Ridiculous $4000 Titan RT Graphics Card?
Facebook may have decided that you shouldn’t see the news, but we think you deserve to be in the know with Gizmodo Australia’s reporting. To sign up for our daily newsletter covering the latest news, features and reviews, head HERE. For a running feed of all our stories, follow us on Twitter HERE. Or you can bookmark the Gizmodo Australia homepage to visit whenever you need a news fix.

For those that have been balking at the $1900 price tag of the RTX 2080 Ti cards, you might want to look away now, as Nvidia has now topped that with the new $3,999 Titan RTX.

Positioned as the follow up to last generation’s Titan V, Nvidia’s new “fastest-ever” PC GPU boasts the company’s new Turing architecture, 576 tensor cores, and 24GB of GDDR6 memory into what should be a ray-tracing powerhouse. So how does it stack up against the rest of Nvidia’s current graphics card lineup?

Specs and comparisons

Nicknamed the T-Rex, when compared to a $1,900 RTX 2080 TI (Nvidia’s previous top-of-the-line consumer GPU), the Titan RTX features slightly more general purpose CUDA graphics cores (4,608 vs 4,352), tensor cores (576 vs 544), and RT cores (72 vs 68)—more on what all that means in a second. And while those increases might not sound like a huge boost for with a $US1,300 ($1,765) premium, the Titan RTX’s specs are almost identical to Nvidia’s powerful workstation-class Quadro RTX 6000, which costs nearly triple the price at over $11,000. So for $3,999, that’s not a bad deal.

Meanwhile, when it comes to memory, the biggest upgrade on the Titan RTX over a regular RTX 2080 Ti is the presence of 24GB of GDDR6 video RAM, which is more than double what you’d get on a 2080 TI.

OK, big numbers are nice, but who is this thing for?

In case you haven’t noticed the general lack of games with support for ray-tracing and DLSS, the Titan RTX’s debut at NeurIPS (Conference on Neural Information Processing Systems) should give you a clue that this thing wasn’t really designed purely with gaming in mind.

By adding even more RT cores for ray tracing and Tensor cores for processing AI and machine learning data, Nvidia’s pricey new card was designed more to help researchers train and develop neural networks faster than any previous GPUs. Additionally, thanks to the Titan RTX’s 100 GBps NV link bridge, scientists can even link two cards to boost performance even further, while the card’s built-in in VirtualLink port provides the kind of video bandwidth needed to test and create next-gen VR experiences.

All these things should work together to help improve things like the AI-generated graphics demo Nvidia also showed off today, which used data gleaned from video footage captured in the real world to generate an entire 3D virtual environment in Unreal Engine 4.

For the regular buyer, this makes the Titan RTX more of a prosumer-level graphics card with a heavy emphasis on data and AI processing. However, if you’re the type of gamer that can’t stand not having the fastest components out there, it’s possible the Titan RTX could deliver 10 to 20 per cent better performance than what you’d get from aanRTX 2080 Ti. Though of course, FPS numbers are sure to vary quite a bit from game to game.

So when it the Titan RTX coming out?

OK, wallet be damned, if you still want a Titan RTX, while Nvidia hasn’t provided an exact release date, the company did say that its new $3,999 GPU will be available sometime before the end of the year in both the U.S. and Europe.

As for Australia, you can get an availability notification from NVIDIA over at the local website.