“This talk will introduce these three debugging techniques and provide some suggestions on selecting the optimal approach for a variety of debugging scenarios such as hangs, numerical errors, and crashes. Specific examples will be given using the TotalView debugger but the concepts covered may apply to other debugging tools such as GDB and the NVIDIA NSIGHT debugger.”
“Learn about extensions that enable efficient use of Partitioned Global Address Space (PGAS) Models like OpenSHMEM and UPC on supercomputing clusters with NVIDIA GPUs. PGAS models are gaining attention for providing shared memory abstractions that make it easy to develop applications with dynamic and irregular communication patterns. However, the existing UPC and OpenSHMEM standards do not allow communication calls to be made directly on GPU device memory. This talk discusses simple extensions to the OpenSHMEM and UPC models to address this issue.”
“We present a state-of-the-art image recognition system, Deep Image, developed using end-to-end deep learning. The key components are a custom-built supercomputer dedicated to deep learning, a highly optimized parallel algorithm using new strategies for data partitioning and communication, larger deep neural network models, novel data augmentation approaches, and usage of multi-scale high-resolution images.”
“We present results for a platform consisting of an NVM Express SSD, a CAPI accelerator card and a software stack running on a Power8 system. We show how the threading of the Power8 CPU can be used to move data from the SSD to the CAPI card at very high speeds and implement accelerator functions inside the CAPI card that can process the data at these speeds.”
Learn how OpenACC runtimes also exposes performance-related information revealing where your OpenACC applications are wasting clock cycles. The talk will show that profilers can connect with OpenACC applications to record how much time is spent in OpenACC regions and what device activity it turns into.
“In this session we describe how GPUs can be used within virtual environments with near-native performance. We begin by showing GPU performance across four hypervisors: VMWare ESXi, KVM, Xen, and LXC. After showing that performance characteristics of each platform, we extend the results to the multi-node case with nodes interconnected by QDR InfiniBand. We demonstrate multi-node GPU performance using GPUDirect-enabled MPI, achieving efficiencies of 97-99% of a non-virtualized system.”