NVIDIA officially announces TITAN RTX with Turing architecture
The Titan guarantees the most demanding users 130 Teraflops of power for the Deep Learning and 11 GigaRay for the performances in Ray Tracing.
The new TITAN RTX is a product intended primarily for researchers and developers of Artificial Intelligence. NVIDIA officially revealed it at the ” Conference on Neural Information Processing Systems ” in Montreal, dubbing it T-Rex precisely to highlight the new acronym, also present on the consumer counterpart, aimed at underlining the hardware implementation of Ray Tracing.
TITAN RTX provides multi-precision Tensor Core for remarkable performance in FP32, FP16, INT8 and INT4, enabling faster training and inference of neural networks. It offers twice the storage capacity of the previous generation TITAN GPUs, together with NVLink to allow researchers to conduct experiments with neural networks and larger data sets.
TITAN RTX accelerates data analysis with NVIDIA RAPIDS, a tool for data scientists. RAPIDS open source libraries integrate seamlessly with the most popular data science workflows in the world to accelerate machine learning.
” TITAN RTX also offers computing power and memory bandwidth required for real-time 8K video editing, ” says NVIDIA. Together with the artificial intelligence engine, the RT Core, along with increased processing power and advanced rasterization, enable functionality that aims to improve the work of millions of developers, designers and artists in various industries.
The video card equips the 13-phase iMON DrMOS power supply unit and is equipped with a steam chamber with a complete board to maximize cooling efficiency.
These are the main data of the new GPU:
- 576 Turing Tensor multi-precision cores, which provide up to 130 teraflops of performance in Deep Learning.
- 72 Turing RT Core, which provide up to 11 GigaRay performance per second Ray Tracing in real time.
- 24 GB of high-speed GDDR6 memory with 672 GB / s of bandwidth – 2 times the memory of the previous generation TITAN GPUs – to adapt to larger models and datasets.
- NVIDIA NVLink 100GB/s can pair two TITAN RTX GPUs to scale memory and compute capacity.
- OS Certification : Windows 7 (64 bit), Windows 10 (64 bit) (April 2018 Update or...
- 4608 NVIDIA CUDA cores running at 1770 MegaHertZ boost clock; NVIDIA Turing...
- New 72 RT cores for acceleration of ray tracing
Product prices and availability are subject to change. Any price and availablility information displayed on Amazon at the time of purchase will apply to the purchase of any products.
