At the 2018 NeurIPS conference in Montréal, Nvidia today announced the Titan RTX. Its new flagship consumer card is capable of delivering 130 TFLOPs of compute performance and 11GigaRays/s of ray-tracing performance. It’s not clear how much RTX-OPS that translates into, which is a term Nvidia coined with the release of its RTX 2XXX series cards.

Just like its predecessors, the Titan RTX is aimed at users who need that tier of performance, rather than just gamers, as noted by Nvidia CEO Jensen Huang:

“Turing is NVIDIA’s biggest advance in a decade – fusing shaders, ray tracing, and deep learning to reinvent the GPU. The introduction of T-Rex puts Turing within reach of millions of the most demanding PC users — developers, scientists and content creators.”

“Turing is NVIDIA’s biggest advance in a decade – fusing shaders, ray tracing, and deep learning to reinvent the GPU. The introduction of T-Rex puts Turing within reach of millions of the most demanding PC users — developers, scientists and content creators.”

There’s 24GB of GDDR6 memory with 672GB/s of total memory bandwidth to process anything those professionals might throw at this card, along with support for the 100GB/s NVLink standard, enabling them to use multiple of these cards together.

Additionally, just like the RTX 2XXX series cards, a USB Type-C port is present on the card itself, allowing users to plug in VR headsets that comply with the brand-new VirtualLink standard. Though, there are no headsets that comply with that standard available for purchase to consumers just yet.

Here are all the features that Nvidia highlights in its announcement:

  • 576 multi-precision Turing Tensor Cores, providing up to 130 teraflops of deep learning performance.
  • 72 Turing RT Cores, delivering up to 11 GigaRays per second of real-time ray-tracing performance.
  • 24GB of high-speed GDDR6 memory with 672GB/s of bandwidth — 2x the memory of previous-generation TITAN GPUs — to fit larger models and datasets.
  • 100GB/s NVIDIA NVLink® can pair two TITAN RTX GPUs to scale memory and compute.
  • Incredible performance and memory bandwidth for real-time 8K video editing.
  • VirtualLink™ port provides the performance and connectivity required by next-gen VR headsets.

As for specifications, there are 6 GPCs, 36 TPCs, 72 SMs, 4608 CUDA cores, and 72 RT Cores. 1350MHz is the base clock, with the boost clock going as high as 1770MHz. The memory is clocked at 7000MHz and offers a data rate of 14Gbps. TDP stands at 280 watts with the card requiring two 8-pin connectors. If you wish to read the specs in more detail, they are available here.

Titan RTX looks very much alike the Titan V when it comes to design, featuring a gold-colored aluminum shroud and backplate. Vapor chamber cooling makes a comeback, as well.

Here’s the most interesting bit: Titan RTX is expected to launch later this month in the U.S. and Europe at the price of $2,499. It’s less than the $3,000 Titan V which was released around the same time last year, but more than twice than the RTX 2080Ti, which released in August earlier this year. Certainly, this card isn’t meant for those who aren’t absolutely sure they need it.

NEXT