Search NASA⌕ Search

SEARCH · Search NASA

Results for “High Performance Networking”

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.

At least 307 records · Page 17
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Stable parallel training of Wasserstein conditional generative adversarial neural networks

In this work, we propose a stable, parallel approach to train Wasserstein conditional generative adversarial neural networks (W-CGANs) under the constraint of a fixed computational budget. Differently from previous distributed GANs training techniques, our approach avoids inter-process communications, reduces the risk of mode collapse and enhances scalability by using multiple generators, each one of them concurrently trained on a single data label. The use of the Wasserstein metric also reduces the risk of cycling by stabilizing the training of each generator. We illustrate the approach on the CIFAR10, CIFAR100, and ImageNet1k datasets, three standard benchmark image datasets, maintaining the original resolution of the images for each dataset. Performance is assessed in terms of scalability and final accuracy within a limited fixed computational time and computational resources. To measure accuracy, we use the inception score, the Fréchet inception distance, and image quality. An improvement in inception score and Fréchet inception distance is shown in comparison to previous results obtained by performing the parallel approach on deep convolutional conditional generative adversarial neural networks as well as an improvement of image quality of the new images created by the GANs approach. Weak scaling is attained on both datasets using up to 2000 NVIDIA V100 GPUs on the OLCF supercomputer Summit.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

O3BNN-R: An Out-Of-Order Architecture for High-Performance and Regularized BNN Inference

Binarized Neural Networks (BNN) have drawn tremendous attention due to significantly reduced computational complexity and memory demand. They have especially shown great potential in cost- and power-restricted domains, such as IoT and smart edge-devices, where reaching a certain accuracy bar is often sufficient, and real-time is highly desired.In this work, we demonstrate that the highly-condensed BNN model can be shrunk significantly further by dynamically pruning irregular redundant edges. Based on two new observations on BNN-specific properties, an out-of-order (OoO) architecture – O3BNN-R, can curtail edge evaluation in cases where the binary output of a neuron can be determined early. Similar to Instruction-Level-Parallelism(ILP), these fine-grained, irregular, runtime pruning opportunities are traditionally presumed to be difficult to exploit. In order to increase the pruning opportunities, we also optimize the training process by adding 2 regularization items in the loss function (1) for pooling pruning and (2) for threshold pruning. We evaluate our design on an FPGA platform using three well-known networks, including VggNet-16, AlexNet for ImageNet, and a VGG-like network for Cifar-10.

Geng, Tong↗
DOE OSTI2018

On Concavity and Utilization Analytics of Wide-Area Network Transport Protocols

Wide-area data transfers in high-performance computing infrastructures and big data applications are increasingly carried over dynamically provisioned dedicated network connections, which provide high capacities with no competing traffic. Throughput of TCP and UDT over wide-area connections depends, often in a complex non-linear way, on the congestion control mechanism, buffer sizes, and the number of parallel streams. In addition, our extensive TCP and UDT throughput measurements and time traces over a suite of physical and emulated 10 Gbps connections with 0-366 ms round-trip times (RTTs) show significant statistical and temporal variations. Consequently, parameter selection and optimization for transport methods require data analytics to complement analytical throughput models. We present analytics based on the concavity-convexity of throughput profiles which provide insights into peak throughput and superior/inferior trend compared to linear interpolations based on RTT. In particular, we propose the utilization-concavity coefficient, a scalar metric that incorporates both utilization and concavity profile to characterize the overall performance of a transport protocol. These measurements-based anaytics enable us to select a transport protocol and its parameters for a given host and connection configuration to achieve high throughput with statistical guarantees.

Rao, Nageswara↗
NASA NTRS1992

Photonics and other approaches to high speed communications

Our research group of 4 faculty and about 10-15 graduate students was actively involved (as a group) in the development of computer communication networks for the last five years. Many of its individuals have been involved in related research for a much longer period. The overall research goal is to extend network performance to higher data rates, to improve protocol performance at most ISO layers and to improve network operational performance. We briefly state our research goals, then discuss the research accomplishments and direct your attention to attached and/or published papers which cover the following topics: scalable parallel communications; high performance interconnection between high data rate networks; and a simple, effective media access protocol system for integrated, high data rate networks.

Maly, Kurt↗
DOE OSTI2021

Record-Low Thermal Boundary Resistance between Diamond and GaN-on-SiC for Enabling Radiofrequency Device Cooling

The implementation of 5G-and-beyond networks requires faster, high-performance, and power-efficient semiconductor devices, which are only possible with materials that can support higher frequencies. Gallium nitride (GaN) power amplifiers are essential for 5G-and-beyond technologies since they provide the desired combination of high frequency and high power. These applications along with terrestrial hub and backhaul communications at high power output can present severe heat removal challenges. The cooling of GaN devices with diamond as the heat spreader has gained significant momentum since device self-heating limits GaN’s performance. However, one of the significant challenges in integrating polycrystalline diamond on GaN devices is maintaining the device performance while achieving a low diamond/GaN channel thermal boundary resistance. In this study, we achieved a record-low thermal boundary resistance of around 3.1 ± 0.7 m 2 K/GW at the diamond/Si 3 N 4 /GaN interface, which is the closest to theoretical prediction to date. The diamond was integrated within ~1 nm of the GaN channel layer without degrading the channel’s electrical behavior. Furthermore, we successfully minimized the residual stress in the diamond layer, enabling more isotropic polycrystalline diamond growth on GaN with thicknesses >2 μm and a ~1.9 μm lateral grain size. More isotropic grains can spread the heat in both vertical and lateral directions efficiently. Using transient thermoreflectance, the thermal conductivity of the grains was measured to be 638 ± 48 W/m K, which when combined with the record-low thermal boundary resistance makes it a leading-edge achievement.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

HydraGNN_OPF_GFM_2026 - Ensemble of predictive graph foundation models for power grid applications

This dataset supports research on graph foundation models for optimal power flow (OPF) on electric grids using HydraGNN. It contains heterogeneous graph representations of PGLib-OPF cases spanning systems from 14 to 13,659 buses, together with packed HDF5 datasets for pretraining, feasibility classification, and N-1 contingency analysis. The release includes OPF solution data, downstream fine-tuning datasets, pretrained HeteroSAGE and HeteroHEAT model checkpoints, hyperparameter-optimization summaries across multiple heterogeneous GNN architectures, and aggregated fine-tuning results for sample-efficiency studies. The dataset is designed to enable scalable training, evaluation, and transfer-learning studies for OPF surrogate modeling, including node-level AC-OPF solution prediction, graph-level prediction, feasibility classification, operating-condition generalization, and contingency-response tasks.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS1992

Semi-interpenetrating polymer network for tougher and more microcracking resistant high temperature polymers

This invention is a semi-interpenetrating polymer network which includes a high performance thermosetting polyimide having a nadic end group acting as a crosslinking site and a high performance linear thermoplastic polyimide. An improved high temperature matrix resin is provided which is capable of performing at 316 C in air for several hundreds of hours. This resin has significantly improved toughness and microcracking resistance, excellent processability and mechanical performance, and cost effectiveness.

Pater, Ruth H.↗
NASA NTRS1992

Current state and future direction of computer systems at NASA Langley Research Center

Computer systems have advanced at a rate unmatched by any other area of technology. As performance has dramatically increased there has been an equally dramatic reduction in cost. This constant cost performance improvement has precipitated the pervasiveness of computer systems into virtually all areas of technology. This improvement is due primarily to advances in microelectronics. Most people are now convinced that the new generation of supercomputers will be built using a large number (possibly thousands) of high performance microprocessors. Although the spectacular improvements in computer systems have come about because of these hardware advances, there has also been a steady improvement in software techniques. In an effort to understand how these hardware and software advances will effect research at NASA LaRC, the Computer Systems Technical Committee drafted this white paper to examine the current state and possible future directions of computer systems at the Center. This paper discusses selected important areas of computer systems including real-time systems, embedded systems, high performance computing, distributed computing networks, data acquisition systems, artificial intelligence, and visualization.

Rogers, James L.↗
NASA NTRS1995

The Linux operating system: An introduction

Linux is a Unix-like operating system for Intel 386/486/Pentium based IBM-PCs and compatibles. The kernel of this operating system was written from scratch by Linus Torvalds and, although copyrighted by the author, may be freely distributed. A world-wide group has collaborated in developing Linux on the Internet. Linux can run the powerful set of compilers and programming tools of the Free Software Foundation, and XFree86, a port of the X Window System from MIT. Most capabilities associated with high performance workstations, such as networking, shared file systems, electronic mail, TeX, LaTeX, etc. are freely available for Linux. It can thus transform cheap IBM-PC compatible machines into Unix workstations with considerable capabilities. The author explains how Linux may be obtained, installed and networked. He also describes some interesting applications for Linux that are freely available. The enormous consumer market for IBM-PC compatible machines continually drives down prices of CPU chips, memory, hard disks, CDROMs, etc. Linux can convert such machines into powerful workstations that can be used for teaching, research and software development. For professionals who use Unix based workstations at work, Linux permits virtually identical working environments on their personal home machines. For cost conscious educational institutions Linux can create world-class computing environments from cheap, easily maintained, PC clones. Finally, for university students, it provides an essentially cost-free path away from DOS into the world of Unix and X Windows.

Bokhari, Shahid H.↗
NASA NTRS1997

Scientific Visualization in High Speed Network Environments

In several cases, new visualization techniques have vastly increased the researcher's ability to analyze and comprehend data. Similarly, the role of networks in providing an efficient supercomputing environment have become more critical and continue to grow at a faster rate than the increase in the processing capabilities of supercomputers. A close relationship between scientific visualization and high-speed networks in providing an important link to support efficient supercomputing is identified. The two technologies are driven by the increasing complexities and volume of supercomputer data. The interaction of scientific visualization and high-speed networks in a Computational Fluid Dynamics simulation/visualization environment are given. Current capabilities supported by high speed networks, supercomputers, and high-performance graphics workstations at the Numerical Aerodynamic Simulation Facility (NAS) at NASA Ames Research Center are described. Applied research in providing a supercomputer visualization environment to support future computational requirements are summarized.

Vaziri, Arsi↗
DOE OSTI2022

Performance trade-offs in reconfigurable networks for HPC

Designing efficient interconnects to support high-bandwidth and low-latency communication is critical toward realizing high performance computing (HPC) and data center (DC) systems in the exascale era. At extreme computing scales, providing the requisite bandwidth through overprovisioning becomes impractical. These challenges have motivated studies exploring reconfigurable network architectures that can adapt to traffic patterns at runtime using optical circuit switching. Despite the plethora of proposed architectures, surprisingly little is known about the relative performances and trade-offs among different reconfigurable network designs. We aim to bridge this gap by tackling two key issues in reconfigurable network design. First, we study how cost, power consumption, network performance, and scalability vary based on optical circuit switch (OCS) placement in the physical topology. Specifically, we consider two classes of reconfigurable architectures: one that places OCSs between top-of-rack (ToR) switches—ToR-reconfigurable networks (TRNs)—and one that places OCSs between pods of racks—pod-reconfigurable networks (PRNs). Second, we tackle the effects of reconfiguration frequency on network performance. Our results, based on network simulations driven by real HPC and DC workloads, show that while TRNs are optimized for low fan-out communication patterns, they are less suited for carrying high fan-out workloads. PRNs exhibit better overall trade-off, capable of performing comparably to a fully non-blocking fat tree for low fan-out workloads, and significantly outperform TRNs for high fan-out communication patterns.

Teh, Min Yee↗
DOE OSTI2026

Event-Driven Neuromorphic Accelerator (Caspian)

Software repository for the Caspian neuromorphic system. In order to deploy small, low-power spiking neural networks at the edge, there needs to be 1) an FPGA design for an event simulator to evaluate these neural networks as well as 2) a high-performance cycle-accurate simulator for training these spiking neural networks. This software package solves these problems by providing the Caspian simulator and the µCaspian hardware design in SystemVerilog.

Mitchell, JohnParker [Microsoft Corporation, Redmo↗
DOE OSTI2017

Deep Packet/Flow Analysis using GPUs

Deep packet inspection (DPI) faces severe performance challenges in high-speed networks (40/100 GE) as it requires a large amount of raw computing power and high I/O throughputs. Recently, researchers have tentatively used GPUs to address the above issues and boost the performance of DPI. Typically, DPI applications involve highly complex operations in both per-packet and per-flow data level, often in real-time. The parallel architecture of GPUs fits exceptionally well for per-packet network traffic processing. However, for stateful network protocols such as TCP, their data stream need to be reconstructed in a per-flow level to deliver a consistent content analysis. Since the flow-centric operations are naturally antiparallel and often require large memory space for buffering out-of-sequence packets, they can be problematic for GPUs, whose memory is normally limited to several gigabytes. In this work, we present a highly efficient GPU-based deep packet/flow analysis framework. The proposed design includes a purely GPU-implemented flow tracking and TCP stream reassembly. Instead of buffering and waiting for TCP packets to become in sequence, our framework process the packets in batch and uses a deterministic finite automaton (DFA) with prefix-/suffix- tree method to detect patterns across out-of-sequence packets that happen to be located in different batches. In conclusion, evaluation shows that our code can reassemble and forward tens of millions of packets per second and conduct a stateful signature-based deep packet inspection at 55 Gbit/s using an NVIDIA K40 GPU.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Deep Packet/Flow Analysis using GPUs

aDeep packet inspection (DPI) faces severe performance challenges in high-speed networks (40/100 GE) as it requires a large amount of raw computing power and high I/O throughputs. Recently, researchers have tentatively used GPUs to address the above issues and boost the performance of DPI. Typically, DPI applications involve highly complex operations in both per-packet and per-flow data level, often in real-time. The parallel architecture of GPUs fits exceptionally well for per-packet network traffic processing. However, for stateful network protocols such as TCP, their data stream need to be reconstructed in a per-flow level to deliver a consistent content analysis. Since the flow-centric operations are naturally anti-parallel and often require large memory space for buffering out-of-sequence packets, they can be problematic for GPUs, whose memory is normally limited to several gigabytes. In this work, we present a highly efficient GPU-based deep packet/flow analysi s framework. The proposed design includes a purely GPU-implemented flow tracking and TCP stream reassembly. Instead of buffering and waiting for TCP packets to become in sequence, our framework process the packets in batch and uses a deterministic finite automaton (DFA) with prefix-/suffix- tree method to detect patterns across out-of-sequence packets that happen to be located in different batches. Evaluation shows that our code can reassemble and forward tens of millions of packets per second and conduct a stateful signature-based deep packet inspection at 55 Gbit/s using an NVIDIA K40 GPU.

Gong, Qian↗
DOE OSTI2015

LADS: Optimizing Data Transfers using Layout-Aware Data Scheduling

While future terabit networks hold the promise of signifi- cantly improving big-data motion among geographically distributed data centers, significant challenges must be overcome even on today s 100 gigabit networks to real- ize end-to-end performance. Multiple bottlenecks exist along the end-to-end path from source to sink. Data stor- age infrastructure at both the source and sink and its in- terplay with the wide-area network are increasingly the bottleneck to achieving high performance. In this paper, we identify the issues that lead to congestion on the path of an end-to-end data transfer in the terabit network en- vironment, and we present a new bulk data movement framework called LADS for terabit networks. LADS ex- ploits the underlying storage layout at each endpoint to maximize throughput without negatively impacting the performance of shared storage resources for other users. LADS also uses the Common Communication Interface (CCI) in lieu of the sockets interface to use zero-copy, OS-bypass hardware when available. It can further im- prove data transfer performance under congestion on the end systems using buffering at the source using flash storage. With our evaluations, we show that LADS can avoid congested storage elements within the shared stor- age resource, improving I/O bandwidth, and data transfer rates across the high speed networks.

Data Management↗