NVIDIA's new TITAN V features the GV100 GPU that packs 12GB of HBM2 - the first TITAN card to feature HBM2 memory technology. "We broke new ground with its new processor architecture, instructions, numerical formats, memory architecture and processor links", said Huang. This comes courtesy of a major redesign of the streaming multiprocessor at the center of the GPU, which should deliver a boost in performance within the same power envelope of its predecessor, the Titan Xp.
Promising to turn a PC into an AI supercomputer, the Titan V is built on Nvidia's Volta GPU architecture and has nine times the raw horsepower of its predecessor. With independent parallel integer and floating-point data paths, Volta is also much more efficient on workloads with a mix of computation and addressing calculations. Additionally, combining the L1 data cache and shared memory unit should improve performance while also simplifying programming. The latest release from this lineup is Titan V GPU, which is created to bring tons of power to the people who are interested in dealing with machine learning applications. Speaking of insanity, Nvidia has mentioned that the Titan V will come at a ludicrous price of US$2,999 (about RM12,222). This container registry includes NVIDIA-optimized deep learning frameworks, third-party managed HPC applications, NVIDIA HPC visualization tools and the NVIDIA TensorRT™ inferencing optimizer.
The Titan V is available today and is limited to two per customer. More information at http://nvidianews.nvidia.com/.
Inside, the new TITAN V features 640 Tensor Cores that are turned for deep learning performance. Copies of reports filed with the SEC are posted on the company's website and are available from NVIDIA without charge.