Designed for deep learning, complex simulation, and torealistic rendering, the PNY Technologies Quadro GV100 Gracs Card can accelerate intense grac and computational workflows. NVIDIA's Volta architecture GPU features 5120 CUDA cores for 14.8 TFLOPS of single precision performance, 7.9 TFLOPS of double precision performance, 29.6 TFLOPS of single precision performance, and with the addition of 640 tensor cores, 118.5 TFLOPS of tensor deep-learning performance. High performance at different precision levels provides acceleration for compute acceleration using the CUDA platform and other APIs including OpenCL, OpenGL, DirectCompute, and more. For vRAM, 32GB of HBM2 memory provides a 4096-bit wide memory bus for 870 GB/s of total bandwidth.