SEARCH · Engineering Papers
Results for “Application Abstraction”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Thin film graded density impactors for high rate off-Hugoniot loading: Application to Ta strength.
Abstract not provided.
A Primer on Bayesian Inference and applications to data analysis.
Abstract not provided.
Attributing Performance Variation from Integrated Application and System Data.
Abstract not provided.
Battery Energy Storage Control: Modeling Uncertainty and Applications (PhD Defense Presentation).
Abstract not provided.
Fatigue Life Improvement of Arrestment Hook Shanks by Application of Laser Peening.
Abstract not provided.
On supervised and unsupervised deep learning applications for materials informatics.
Abstract not provided.
Fatigue Life Improvement of Arrestment Hook Shanks by Application of Laser Peening.
Abstract not provided.
Forecasting multi-wave epidemics through Bayesian Inference - Application to COVID-19.
Abstract not provided.
Engineering and control of advanced optoelectronic materials for applications in quantum devices.
Abstract not provided.
Introduction to Structural Dynamics and its Application to System Design.
Abstract not provided.
Cyber-Physical Security Analysis of a DER Application for IoT Communications.
Abstract not provided.
GPU Direct I/O with HDF5
Exascale HPC systems are being designed with accelerators, such as GPUs, to accelerate parts of applications. In machine learning workloads as well as large-scale simulations that use GPUs as accelerators, the CPU (or host) memory is currently used as a buffer for data transfers between GPU (or device) memory and the file system. If the CPU does not need to operate on the data, then this is sub-optimal because it wastes host memory by reserving space for duplicated data. Furthermore, this “bounce buffer” approach wastes CPU cycles spent on transferring data. A new technique, NVIDIA GPUDirect Storage (GDS), can eliminate the need to use the host memory as a bounce buffer. Thereby, it becomes possible to transfer data directly between the device memory and the file system. This direct data path shortens latency by omitting the extra copy and enables higher-bandwidth. To take full advantage of GDS in existing applications, it is necessary to provide support with existing I/O libraries, such as HDF5 and MPI-IO, which are heavily used in applications. In this paper, we describe our effort of integrating GDS with HDF5, the top I/O library at NERSC and at DOE leadership computing facilities. We design and implement this integration using a HDF5 Virtual File Driver (VFD). The GDS VFD provides a file system abstraction to the application that allows HDF5 applications to perform I/O without the need to move data between CPUs and GPUs explicitly. We compare performance of the HDF5 GDS VFD with explicit data movement approaches and demonstrate superior performance with the GDS method.