NVIDIA has officially unveiled its most powerful desktop GPU till date called the TITAN RTX. The GPU is based on Turing architecture and offers breath taking performance.
The NVIDIA TITAN RTX provides multi-precision Turing Tensor Cores for breakthrough performance from FP32, FP16, INT8 and INT4, allowing faster training and inference of neural networks.
The GPU is not just limited to real-time ray-traced graphics. It offers enhanced performance in AI research, creative apps and data science. There is a VirtualLink port which delivers raw power and connectivity required by next-generation VR headsets.
The NVIDIA TITAN RTX boasts 576 multi-precision Turing Tensor core which provides up to 130 teraflops of deep learning performance. It has 72 Turin RT cores which delivers up to 11 GigaRays per second of real-time ray tracing performance.
Not to mention is 24GB of high-speed GDDR6 memory (14Gbps memory speed and 384-bit memory interface width) which offers 672GB/s of bandwidth. The memory twice that of previous generation TITAN GPU’s. One can also pair two TITAN RTX GPUS thanks to 100GB/s NVIDIA NVLink. Its raw performance and memory bandwidth enables real-time 8K video editing.
It has 4608 CUDA Cores with base clock of 1350MHz and Boos clock of 1770MHz. The Graphic card consumes about 280W of power.
The NVIDIA TITAN RTX price is set at $2499 in US and Europe and will be available later this month.