Thursday, January 12, 2017
3:00 PM - 5:30 PM
The Tesla P100 delivers unprecedented performance for hyperscale and HPC applications. It offers 5.3 TeraFLOPS of peak double-precision performance—3X faster than the previous-generation Tesla K40 GPU. It also delivers 10.6 TeraFLOPS of peak single-precision performance to accelerate applications in machine learning, AI and other HPC workloads. This performance is realized in software applications around the world by developers building their applications using NVIDIA CUDA 8, the latest update to NVIDIA’s powerful parallel computing platform and programming model.
This session will present the Pascal hardware architecture, with an emphasis on how CUDA 8 enabled applications and frameworks are able to take advantage of the latest features found in the Tesla P100 and related GPUs.
Participants should bring a laptop to class with an SSH client already installed. PuTTY is the recommended SSH client.