NVIDIA Introduces the TITAN RTX


If the NVIDIA GeForce RTX 20-series graphics cards were made for powerful gaming performance, then the TITAN RTX GPU was engineered for AI research, data science, and creative applications.

Driven by the Turing architecture, the TITAN RTX features 576 multi-precision Turing Tensor Cores, providing up to 130 TFLOPs of deep learning performance. It also has 72 Turing RT Cores that can deliver up to 11 GigaRays per second of ray-tracing performance.


You also won't have any shortage of dedicated VRAM, with 24GB of GDDR6 and 672 GB/s of bandwidth, doubling the memory of previous-generation TITAN cards. Need even more power? You can pair two TITAN RTX cards using a 100GB/s NVIDIA NVLink to scale memory and compute. You can also use the VirtualLink USB Type-C port for next-gen VR headsets.

The TITAN RTX can help AI researchers and developers provide faster training and inference of neural networks, thanks to its breakthrough performance from FP32, FP16, INT8, and INT4. Pairing two TITAN RTX cards will also allow researchers to experiment with larger neural networks and data sets.


Data scientists can also take advantage of the TITAN RTX by using the card to accelerate data analytics with RAPIDS. The RAPIDS open-source libraries can speed up machine learning by integrating seamlessly with popular data science workflows.

The TITAN RTX isn't just limited to AI research. With the power of real-time ray tracing and AI to creative applications, the TITAN RTX has the bandwidth needed for real-time 8K video editing.


Are you ready for the most powerful desktop GPU to date? Let us know in the Comments section, below.