Search NASA⌕ Search

SEARCH · Search NASA

Results for “bandwidth optimization”

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 109 records · Page 6

Efficient on-chip terahertz generation and detection with GaN photoconductive emitters

Photoconductive emitters for terahertz generation hold promise for highly efficient down-conversion of optical photons because it is not constrained by the Manley-Rowe relation. Existing terahertz photoconductive devices, however, faces limits in efficiency due to the semiconductor properties of commonly used GaAs materials. Here, we demonstrate that large bandgap semiconductor GaN, characterized by its high breakdown electric field, facilitates the highly efficient generation of terahertz waves in a coplanar stripline waveguide. Towards this goal, we investigated the excitonic contribution to the electro-optic response of GaN under static electric field both through experiments and first-principles calculations, revealing a robust excitonic Stark shift. Using this electro-optic effect, we developed a novel ultraviolet pump-probe spectroscopy for in-situ characterization of the terahertz electric field strength generated by the GaN photoconductive emitter. Our findings show that terahertz power scales quadratically with optical excitation power and applied electric field over a broad parameter range. We achieved an optical-to-terahertz conversion efficiency approaching 100% within the 0.03–1 THz bandwidth at the highest bias field (116 kV/cm) in our experiment. Further optimization of GaN-based terahertz generation devices could achieve even greater optical-to-terahertz conversion efficiencies.

Uzundal, Can B↗

A High Performance Block Eigensolver for Nuclear Configuration Interaction Calculations

As on-node parallelism increases and the performance gap between the processor and the memory system widens, achieving high performance in large-scale scientific applications requires an architecture-aware design of algorithms and solvers. We focus on the eigenvalue problem arising in nuclear Configuration Interaction (CI) calculations, where a few extreme eigenpairs of a sparse symmetric matrix are needed. Here, we consider a block iterative eigensolver whose main computational kernels are the multiplication of a sparse matrix with multiple vectors (SpMM), and tall-skinny matrix operations. We then present techniques to significantly improve the SpMM and the transpose operation SpMM T by using the compressed sparse blocks (CSB) format. We achieve 3-4× speedup on the requisite operations over good implementations with the commonly used compressed sparse row (CSR) format. We develop a performance model that allows us to correctly estimate the performance of our SpMM kernel implementations, and we identify cache bandwidth as a potential performance bottleneck beyond DRAM. We also analyze and optimize the performance of LOBPCG kernels (inner product and linear combinations on multiple vectors) and show up to 15× speedup over using high performance BLAS libraries for these operations. The resulting high performance LOBPCG solver achieves 1.4× to 1.8× speedup over the existing Lanczos solver on a series of CI computations on high-end multicore architectures (Intel Xeons). We also analyze the performance of our techniques on an Intel Xeon Phi Knights Corner (KNC) processor.

97 MATHEMATICS AND COMPUTING↗

Evaluation of the Single-precision Floatingpoint Vector Add Kernel Using the Intel FPGA SDK for OpenCL

Open Computing Language (OpenCL) is a high-level language that enables software programmers to explore Field Programmable Gate Arrays (FPGAs) for application acceleration. The Intel FPGA software development kit (SDK) for OpenCL allows a user to specify applications at a high level and explore the performance of low-level hardware acceleration. In this report, we present the FPGA performance and power consumption results of the single-precision floating-point vector add OpenCL kernel using the Intel FPGA SDK for OpenCL on the Nallatech 385A FPGA board. The board features an Arria 10 FPGA. We evaluate the FPGA implementations using the compute unit duplication and kernel vectorization optimization techniques. On the Nallatech 385A FPGA board, the maximum compute kernel bandwidth we achieve is 25.8 GB/s, approximately 76% of the peak memory bandwidth. The power consumption of the FPGA device when running the kernels ranges from 29W to 42W.

97 MATHEMATICS AND COMPUTING↗

Study Results and Plans for Future Improvement of the NSLS-II FOFB System

Currently, the fast orbit feedback (FOFB) system correct the transverse orbit in both planes via air coil magnets installed over low conductivity sections of vacuum chamber (stainless steel bellows) with a fast acquisition (FA) that samples at 10kHz. Bandwidth of a few kHz has been achieved (power-supply + corrector + vacuum chamber) for the purpose that the FOFB bandwidth limitation is not anymore limited by the correctors. The bandwidths of the FOFB system in both planes have been experimentally characterized in the open-loop condition via sinwave excitation method. They are significantly lower than the measurements (2.5kHz in x) done by Weixing Chen et al when the control system was largely bypassed, only from PSI to fast correctors, final to BPM. The estimated dI/dt as a function of the frequency from the beam noise spectrum has been compared to the slew rate of the power supply. The most recent study evidences: after the optimization of fast corrector compensation circuit, the power supply system along shows the 3dB bandwidth of 5.6kHz in the horizontal plane and 3.5kHz in the vertical plane. Preliminary beam studies indicate the potential increase of the system bandwidth and the gain within the FOFB bandwidth. Finally, the plan for combing the FOFB and SOFB systems to correct down to DC has been discussed.

36 MATERIALS SCIENCE↗

Circular Corrugated Miter Bend and Gap Losses for Broadband Frequency Applications

Circular corrugated waveguides are often used in fusion applications at single, multiple, or broadband millimeter frequencies due to their low ohmic loss, expected large frequency bandwidth, and direct coupling to free-space Gaussian modes. For single-frequency corrugated waveguide systems, transmission line components can be optimized to the desired frequency. For broadband or multiple-frequency applications, this is not possible. The goal of this paper is to demonstrate that the frequency bandwidth of circular corrugated waveguides can be compromised by diffraction losses to miter bends and gaps. It is shown that if the corrugation depth differs significantly from λ /4, a theory can substantially underpredict the gap and miter bend losses. The simulations are also shown to compare favorably to experimental measurements. To improve the transmission line performance in large frequency bandwidth systems, such as 33–165 GHz reflectometry, reducing the number of miter bends may, therefore, be necessary. For the improvement of performance in narrower frequency bandwidth or multiple-frequency systems, especially high-powered systems, wavelength-dependent techniques may be applicable. Furthermore, one such application is shown for a Bragg reflection technique to reduce the thermal load to insulating ceramic rings in a dc break component for multiple-frequency ITER electron cyclotron heating transmission lines.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Slow light engineering in a photonic crystal slab waveguide through optofluidic infiltration and geometric modulation

In this paper, a new type of flat-band slow light structure with high group index (n{sub g}) and large normalized delay-bandwidth product (NDBP) in a silicon on insulator (SOI) based photonic crystal (PC) slab waveguide with a triangular lattice of circular holes is demonstrated. The dispersion engineering is performed by infiltrating optical fluids with different refractive indices n{sub f} in the first row and shifting the second row of air holes adjacent to the PC waveguide (PCW) in the longitudinal direction. In the optimized case, a high NDBP of 0.32 with a group index of 54.55 and a bandwidth of 9.13 nm could be obtained. Furthermore, an ultra-low group velocity dispersion (GVD) in the range of 10{sup –20} s{sup 2}/m is achieved in all of the structures. These results are obtained by numerical simulations based on three-dimensional (3D) plane wave expansion (PWE) method.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Broadband laser output by application of stimulated rotational Raman scattering during high-intensity propagation of laser pulses with modified spectral profiles

Research at the Naval Research Laboratory pursues advanced laser technologies that can mitigate laser plasma instabilities (LPI) in applications such as inertial confinement fusion. Increased laser bandwidth is currently being developed as a strategy to suppress LPI. In particular, stimulated rotational Raman scattering (SRRS) has been demonstrated using the Nike krypton fluoride (KrF) laser facility as a viable method for enhancing bandwidth. The technique is applicable to other large laser facilities. Through a combination of high-intensity propagation and optimization of the laser spectrum in the low-energy stages of the laser, the output spectrum of Nike has been broadened with SRRS from its standard operating value of 1 THz full width at half-maximum (FWHM) bandwidth to span a range greater than 4.5 THz. This amount of bandwidth has been shown in simulations to be effective for mitigation of slow-growth laser plasma instabilities, such as cross-beam energy transport (CBET).

Optics↗

VISAR Analysis in the Frequency Domain

VISAR measurements are typically analyzed in the time domain, where velocity is approximately proportional to fringe shift. Moving to the frequency domain clarifies the limitations of this approximation and suggests several improvements. For example, optical dispersion preserves high-frequency information, so a zero-dispersion (air delay) interferometer does not provide optimal time resolution. Combined VISAR measurements can also improve time resolution. With adequate bandwidth and reasonable noise levels, it is quite possible to achieve better resolution than the VISAR approximation allows.

36 MATERIALS SCIENCE↗

High Performance Adaptive Physics Refinement to Enable Large-Scale Tracking of Cancer Cell Trajectory

The ability to track simulated cancer cells through the circulatory system, important for developing a mechanistic understanding of metastatic spread, pushes the limits of today's supercomputers by requiring the simulation of large fluid volumes at cellular-scale resolution. To overcome this challenge, we introduce a new adaptive physics refinement (APR) method that captures cellular-scale interaction across large domains and leverages a hybrid CPU-GPU approach to maximize performance. Through algorithmic advances that integrate multi-physics and multi-resolution models, we establish a finely resolved window with explicitly modeled cells coupled to a coarsely resolved bulk fluid domain. In this work we present multiple validations of the APR framework by comparing against fully resolved fluid-structure interaction methods and employ techniques, such as latency hiding and maximizing memory bandwidth, to effectively utilize heterogeneous node architectures. Collectively, these computational developments and performance optimizations provide a robust and scalable framework to enable system-level simulations of cancer cell transport.

59 BASIC BIOLOGICAL SCIENCES↗

Bandwidth extension and conversion efficiency improvements beyond phase matching limitations using cavity-enhanced OPCPA

The conversion efficiency and phase matching bandwidth of ultrafast optical parametric amplification (OPA) are constrained by the dispersion and nonlinear coefficient of the employed crystal as well as pulse shaping effects. In our work we show that an enhancement cavity resonant with the pump seeded at the full repetition rate of the pump laser can automatically reshape the small-signal gain in optical parametric chirped-pulse amplification (OPCPA) to achieve close-to-optimal operation. This new method termed cavity-enhanced OPCPA or C-OPCPA significantly increases both the gain bandwidth and the conversion efficiency, in addition to boosting gain for high-repetition-rate amplification. The goal in C-OPCPA is to arrive at a condition of impedance matching at all temporal coordinates, such that, in the absence of linear losses, all the incident pump power is dissipated in the nonlinear loss element, i.e., converted to signal and idler. The use of a low finesse enhancement cavity resonant with a low average power (<1W) and a high repetition rate (78MHz) pump source is shown to achieve more than 50% conversion efficiency into signal and idler from the coupled pump in an optical parametric process, whereas an equivalent amount of pump power in a single-pass configuration leads to negligible conversion. Additionally, the gain bandwidth is extended by a factor of 3-4 beyond the phase-matching limit. Our empirical observations are corroborated by a numerical analysis of depletion optimizing the single-pass case, which assesses the underlying impedance matching that is responsible for the observed performance improvements.

47 OTHER INSTRUMENTATION↗

Dispersion Caused by the Penetration Effect in X-ray Compressors

Chirped X-ray pulse compression is a promising approach for generating ultra-short X-ray free electron laser (XFEL) pulses. The design of X-ray pulse compressors requires the careful control of group delay dispersion (GDD), which plays a critical role in achieving optimal compression. However, the penetration dispersion of crystals and multilayers can induce an extra GDD, which may result in over-compression or under-compression. In this study, we investigate the penetration dispersion of crystals and multilayers theoretically and numerically. Our results indicate that the extra GDD induced by the penetration effect increases as the bandwidth of the rocking curve decreases. Moreover, the extra GDD is nonlinear and can be mitigated by optimizing the configuration of X-ray pulse compressors. This work provides insights into the dispersion compensation and configuration optimization of X-ray pulse compressors, which are essential for generating ultra-short XFEL pulses.

47 OTHER INSTRUMENTATION↗

A two-axis in-plane motion measurement system based on optical beam deflection

Measurement of in-plane motion with high resolution and large bandwidth enables model-identification and real-time control of motion-stages. This paper presents an optical beam deflection based system for measurement of in-plane motion of both macro- and micro-scale motion stages. A curved reflector is integrated with the motion stage to achieve sensitivity to in-plane translational motion along two axes. Under optimal settings, the measurement system is shown to theoretically achieve sub-angstrom measurement resolution over a bandwidth in excess of 1 kHz and negligible cross-sensitivity to linear motion. Subsequently, the proposed technique is experimentally demonstrated by measuring the in-plane motion of a piezo flexure stage and a scanning probe microcantilever. For the former case, reflective spherical balls of different radii are employed to measure the in-plane motion and the measured sensitivities are shown to agree with theoretical values, on average, to within 8.3%. For the latter case, a prototype polydimethylsiloxane micro-reflector is integrated with the microcantilever. The measured in-plane motion of the microcantilever probe is used to identify nonlinearities and the transient dynamics of the piezo-stage upon which the probe is mounted. These are subsequently compensated by means of feedback control.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Minimum Spectral Bandwidth in Echo Seeded Free Electron Lasers

This paper examines the impact of non-linear longitudinal phase distortions on the spectral bandwidth in echo seeded free electron lasers (FELs). It extends the existing theory developed in Hemsing for echo-enabled harmonic generation (EEHG) to include finite laser pulse durations. An analytic expression for the shape of the optimized longitudinal bunching envelope is derived, and is used to determine the laser and electron beam pulse durations that minimize the seeded bandwidth in the presence of arbitrary phase distortions. The time-bandwidth product (TBP) is also derived, and is shown that the TBP and the bandwidth increase by no more than 2–√ from their transform-limited values when the bandwidth is minimized.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Minimum Spectral Bandwidth in Echo Seeded Free Electron Lasers

This paper examines the impact of non-linear longitudinal phase distortions on the spectral bandwidth in echo seeded free electron lasers (FELs). It extends the existing theory developed in Ref.~\cite{HemsingBunchingPhase} for echo-enabled harmonic generation (EEHG) to include finite laser pulse durations. An analytic expression for the shape of the optimized longitudinal bunching envelope is derived, and is used to determine the laser and electron beam pulse durations that minimize the seeded bandwidth in the presence of arbitrary phase distortions. The time-bandwidth product (TBP) is also derived, and is shown that the TBP and the bandwidth increase by no more than $\sqrt{2}$ from their transform-limited values when the bandwidth is minimized.

Hemsing, Erik↗

Efficient GPU Implementation of Automatic Differentiation for Computational Fluid Dynamics

Many scientific and engineering applications require repeated calculations of derivatives of output functions with respect to input parameters. Automatic Differentiation (AD) is a method that automates derivative calculations and can significantly speed up code development. In Computational Fluid Dynamics (CFD), derivatives of flux functions with respect to state variables (Jacobian) are needed for efficient solutions of the nonlinear governing equations. AD of flux functions on graphics processing units (GPUs) is challenging as flux computations involve many intermediate variables that create high register pressure and require significant memory traffic because of the need to store the derivatives. This paper presents a forward-mode AD method based on multivariate dual numbers that addresses these challenges and simultaneously reduces the floating-point operation count. The dimension of the multivariate dual numbers is optimized for performance. The flux computations are restructured to minimize the number of temporary variables and reduce register pressure. For effective utilization of memory bandwidth, shared memory is used to store the local flux Jacobian. This AD implementation is compared with several other Jacobian implementations on an NVIDIA V100 GPU (V100). For three-dimensional perfect-gas compressible-flow equations implemented in a practical CFD code, the AD implementation of a flux Jacobian based on multivariate dual numbers of dimension 5 outperforms all other GPU AD implementations on V100. Its performance is comparable with the optimized hand-differentiated version. Finally, the implementation achieves 75% of the peak floating-point throughput and 61 % of the peak global device memory bandwidth usage.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Stochastic optimization of GeantV code by use of genetic algorithms

GeantV is a complex system based on the interaction of different modules needed for detector simulation, which include transport of particles in fields, physics models simulating their interactions with matter and a geometrical modeler library for describing the detector and locating the particles and computing the path length to the current volume boundary. The GeantV project is recasting the classical simulation approach to get maximum benefit from SIMD/MIMD computational architectures and highly massive parallel systems. This involves finding the appropriate balance between several aspects influencing computational performance (floating-point performance, usage of off-chip memory bandwidth, specification of cache hierarchy, etc.) and handling a large number of program parameters that have to be optimized to achieve the best simulation throughput. This optimization task can be treated as a black-box optimization problem, which requires searching the optimum set of parameters using only point-wise function evaluations. Here, the goal of this study is to provide a mechanism for optimizing complex systems (high energy physics particle transport simulations) with the help of genetic algorithms and evolution strategies as tuning procedures for massive parallel simulations. One of the described approaches is based on introducing a specific multivariate analysis operator that could be used in case of resource expensive or time consuming evaluations of fitness functions, in order to speed-up the convergence of the black-box optimization problem.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Deep Generative Models that Solve PDEs: Distributed Computing for Training Large Data-Free Models

Recent progress in scientific machine learning (SciML) has opened up the possibility of training novel neural network architectures that solve complex partial differential equations (PDEs). Several (nearly data free) approaches have been recently reported that successfully solve PDEs, with examples including deep feed forward networks, generative networks, and deep encoder-decoder networks. However, practical adoption of these approaches is limited by the difficulty in training these models, especially to make predictions at large output resolutions (≥1024×1024). Here we report on a software framework for data parallel distributed deep learning that resolves the twin challenges of training these large SciML models - training in reasonable time as well as distributing the storage requirements. Our framework provides several out of the box functionality including (a) loss integrity independent of number of processes, (b) synchronized batch normalization, and (c) distributed higher-order optimization methods. We show excellent scalability of this framework on both cloud as well as HPC clusters, and report on the interplay between bandwidth, network topology and bare metal vs cloud. We deploy this approach to train generative models of sizes hitherto not possible, showing that neural PDE solvers can be viably trained for practical applications. We also demonstrate that distributed higher-order optimization methods are 2-3× faster than stochastic gradient-based methods and provide minimal convergence drift with higher batch-size.

PDEs↗