Following a rash of strategic 'leaks' by social media influencers, NVIDIA on Monday formally introduced its Titan RTX graphics card, a heavy-hitting accelerator that the company is appropriately billing as "the world's most powerful desktop GPU." It's built to handle the data-crunching rigors of deep learning applications, and of course brings ray tracing performance to the fold.
As with all RTX card, the Titan RTX—or T-Rex, as NVIDIA has nicknamed it—is built around the company's Turing GPU. NVIDIA claims its newest graphics card can pummel workloads with 130 teraflops of deep learning performance and 11 gigarays or ray tracing performance.
"Turing is NVIDIA’s biggest advance in a decade—fusing shaders, ray tracing, and deep learning to reinvent the GPU," said Jensen Huang, founder and CEO of NVIDIA. "The introduction of T-Rex puts Turing within reach of millions of the most demanding PC users—developers, scientists and content creators."
It was rumored the Titan RTX might 'only' ship with 12GB of GDDR6 memory, but is instead outfitted with twice that amount—T-Rex wields 24GB of GDDR6 memory underneath the hood, pushing 672GB/s of memory bandwidth. For comparison, the GeForce RTX 2080 Ti wields 11GB of GDDR6 memory.
Titan RTX is essentially an unabashed representation of the full Turing TU102 GPU. It has 4,608 CUDA cores, 576 Tensor cores, and 72 RT cores. Clockspeeds check in at 1,350MHz (base) and 1,770MHz (boost). Users with deep pockets can pair two Titan RTX cards to scale memory and compute performance via NVLink.
This is a dual-slot card with the same retooled cooler design as found on NVIDIA's GeForce RTX Founders Edition models. It draws power from a pair of 8-pin PCIe connectors, with a 280W TDP (NVIDIA recommends a 650W PSU or higher). Connectors consist of three DisplayPorts, a single HDMI port, and a USB-Type C connector.
NVIDIA says the Titan RTX will be available later this month in the US and Europe for $2,499.