NVIDIA Turing GPU Architecture
Armed with the all-new RTCore for ray tracing, 384 Tensor Cores for AI and 3072 CUDA cores for parallel computing, NVIDIA Turing is simply the world's most advanced GPU.
Equipped with 16GB of ultra-fast GDDR6 memory to hold large datasets – complex designs for products, architectural walkthroughs, media assets and more.
Link two GPUs with a high-speed interconnect to scale memory capacity to 32GB and drive higher performance with up to 50 GB/s of data transfer.
Be ready for next-generation of high-resolution VR head-mounted displays and enjoy simplified cabling with support for the industry standard VirtualLink connector.