Search NASA⌕ Search

SEARCH · Search NASA

Results for “Mixed precision”

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 127 records · Page 7

Performance Optimization Methods for a Memory-Bound, Unstructured-Grid CFD Application on Massively Parallel GPU Platforms

Computational performance of the FUN3D unstructured-grid computational fluid dynamics (CFD) application on massively parallel GPU environments is memory-bound and highly dependent upon efficient reads from and atomic updates to the irregular cell-, edge-, and node-based data structures. In this talk, we present recent efforts into optimizing select performance-critical kernels on NVIDIA Tesla V100 and A100 GPUs and AMD CDNA MI100 GPUs. A novel use of L2 cache residency controls and asynchronous loads into on-chip shared memory are explored on the A100 GPU for the sparse iterative solver, which is dominated by mixed-precision, sparse matrix vector multiplication. Demonstrations show that these methods improve global memory bandwidth utilization by 13.5% on the A100 GPU. Several techniques are also presented that use registers and/or shared memory to facilitate array transposition and aggregation which combine to reduce the frequency and increase the cache efficiency of floating-point atomic updates to the irregular data structures. These methods are demonstrated to improve the kernel throughput by nearly 500% on select kernels on the AMD MI100 over atomic updates directly to global memory. Overall, both V100 and A100 GPUs outperformed the MI100 GPU on kernels dominated by double-precision atomic updates; however, the techniques demonstrated here reduced the performance gap and improved the MI100 performance.

GPU CPU unstructured CFD memory↗

xSDK: Building an ecosystem of highly efficient math libraries for exascale

Current efforts to build increasingly powerful computer architectures are opening up new avenues for more complex and higher fidelity simulations coupled with data analytics and learning, leading to new scientific insights and deeper understanding. At one extreme, exascale computers will be much faster than previous computer generations (performing 10 18 operations per second—that is, 1,000 times faster than petascale). To achieve these performance improvements, computer architectures are becoming increasingly complex, with deep memory hierarchies, very high node and core counts, and heterogeneous features such as graphics processing units (GPUs). Such architectural changes impact the full breadth of computing scales, as heterogeneity pervades even current-generation laptops, workstations, and moderate-sized clusters. While emerging advanced architectures provide unprecedented opportunities, they also present significant challenges for developers of scientific applications, such as multiphysics and multiscale codes, who must adapt their software to handle disruptive changes in architectures and new programming models that have not yet stabilized. Developers must consider increasing concurrency while reducing communication and synchronization, and other complexities such as the potential for using mixed precision to leverage the compute power available in low-precision tensor cores. On one hand, developers must implement new scientific capabilities, which in turn increase code complexity. On the other hand, the codes must be ported to new architectures, requiring the inclusion of new programming models and the restructuring of code to achieve good performance. Addressing these issues is beyond the capability of any single person or team—leading to the need for collaboration among many teams, who encapsulate their expertise in reusable software and work together to create sustainable software ecosystems.

97 MATHEMATICS AND COMPUTING↗

Using Low Energy Atmospheric Neutrinos for Precision Measurement of the Mixing Parameters

Use of low energy atmospheric neutrinos is considered for precision measurement of neutrino mixing parameters. At around energy $E \simeq$ a few $\times$100 MeV and baseline $L$ of a few $\times$1000 km, CP phase effect is $\sim$10 times larger than that of the conventional LBL accelerator neutrino experiments. We report here a few progresses: (1) To analyze physics in the region, a new perturbative framework at around the solar-scale enhancement is developed. (2) To know the characteristic features of CP $\delta$ dependence of the atmospheric neutrinos at low energies, we plot the ratio of the $\nu_{e}$ and $\bar{\nu}_{e}$ fluxes $F(\delta)/F(\delta=0)$ for $\delta=\pm \frac{\pi}{2}$ and $\pi$ as a function of $E$ using the Honda {\it et al.} flux. Interestingly, it shows $\delta$ dependence of $\simeq$5-10\% level, with positive (negative) sign for $\nu_{e}$ ($\bar{\nu}_{e}$). (3) To reduce the flux systematic errors, measurement of muon energy distribution around $\sim$1 GeV at high altitude may be useful. Water Cherenkov or muon range detectors may be the options with advantage in the latter if it is magnetized in view of the $\delta$ dependence of the flux.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Results of Propellant Mixing Variable Study Using Precise Pressure-Based Burn Rate Calculations

A designed experiment was conducted in which three mix processing variables (pre-curative addition mix temperature, pre-curative addition mixing time, and mixer speed) were varied to estimate their effects on within-mix propellant burn rate variability. The chosen discriminator for the experiment was the 2-inch diameter by 4-inch long (2x4) Center-Perforated (CP) ballistic evaluation motor. Motor nozzle throat diameters were sized to produce a common targeted chamber pressure. Initial data analysis did not show a statistically significant effect. Because propellant burn rate must be directly related to chamber pressure, a method was developed that showed statistically significant effects on chamber pressure (either maximum or average) by adjustments to the process settings. Burn rates were calculated from chamber pressures and these were then normalized to a common pressure for comparative purposes. The pressure-based method of burn rate determination showed significant reduction in error when compared to results obtained from the Brooks' modification of the propellant web-bisector burn rate determination method. Analysis of effects using burn rates calculated by the pressure-based method showed a significant correlation of within-mix burn rate dispersion to mixing duration and the quadratic of mixing duration. The findings were confirmed in a series of mixes that examined the effects of mixing time on burn rate variation, which yielded the same results.

Stefanski, Philip L.↗

Indirect precise angular control using four-wave mixing

Here, we show indirect precise angular control using a four-wave mixing (FWM) process. This was performed with a superposition of light with orbital angular momentum in an M-Type configuration of a hot {sup 85}Rb atomic ensemble. A gear-shaped interference pattern is observed at FWM light with a donut-shaped input signal. The gear could be rotated and is controlled through the change of the polarization of the pump laser. Our experimental results that are based on nonlinear coherent interactions have applications in image processing and precise angular control.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The P2 experiment: A future high-precision measurement of the weak mixing angle at low momentum transfer

Here we describe the research and development work for the P 2 experiment which aims for a high precision determination of the weak mixing angle sin 2 θ W to a precision of 0.15% at a four-momentum transfer of 4.5×10 –3 GeV 2 . This accuracy, comparable to existing measurements at the Z pole, allows for a sensitive test of the Standard Model up to a mass scale of 50 TeV, extendable to 60 TeV. The weak mixing angle is connected to the weak charge of the proton which will be extracted from a measurement of the parity violating cross section asymmetry –39.94×10 –9 in elastic electron-proton scattering. A total accuracy of 0.57×10 –9 is achievable in a measurement time of 11000 h using a 150μA polarized electron beam impinging on a 60 cm liquid hydrogen target. The P 2 asymmetry is smaller than any asymmetry measured so far in electron scattering with an unprecedented goal for the accuracy. The use of a solenoid spectrometer with 100% Φ-acceptance as well as an atomic hydrogen trap polarimeter are new features, which have never before been used in parity-violation experiments. In order to collect the enormous statistics required for this measurement, the new Mainz Energy-Recovering Superconducting Accelerator (MESA) is under construction. Plans for the associated beam control system and the polarimetry are described in this article as well. A liquid hydrogen high-power target with an extremely low noise level of 10 ppm needs to be designed and constructed. We report in addition on the conceptual design of the P 2 spectrometer, its Cherenkov detectors, the integrating read-out electronics as well as the ultra-thin, fast tracking detectors. The physics program of the MESA facility comprises indirect, high precision search for physics beyond the Standard Model, measurement of the neutron distribution in nuclear physics, single-spin asymmetries, and a possible future extension to the measurement of hadronic parity violation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

The MOLLER Experiment: An Ultra-Precise Measurement of the Weak Mixing Angle Using Møller Scattering

The physics case and an experimental overview of the MOLLER (Measurement Of a Lepton Lepton Electroweak Reaction) experiment at the 12 GeV upgraded Jefferson Lab are presented. A highlight of the Fundamental Symmetries subfield of the 2007 NSAC Long Range Plan was the SLAC E158 measurement of the parity-violating asymmetry $A_{PV}$ in polarized electron-electron (M{\o}ller) scattering. The proposed MOLLER experiment will improve on this result by a factor of five, yielding the most precise measurement of the weak mixing angle at low or high energy anticipated over the next decade. This new result would be sensitive to the interference of the electromagnetic amplitude with new neutral current amplitudes as weak as $\sim 10^{-3}\cdot G_F$ from as yet undiscovered dynamics beyond the Standard Model. The resulting discovery reach is unmatched by any proposed experiment measuring a flavor- and CP-conserving process over the next decade, and yields a unique window to new physics at MeV and multi-TeV scales, complementary to direct searches at high energy colliders such as the Large Hadron Collider (LHC). The experiment takes advantage of the unique opportunity provided by the upgraded electron beam energy, luminosity, and stability at Jefferson Laboratory and the extensive experience accumulated in the community after a round of recent successfully completed parity-violating electron scattering experiments

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Conflict between the identification of a cosmic neutrino source and the sensitivity to mixing angles in a neutrino telescope

Neutrino fluxes at telescopes depend on both initial fluxes out of astronomical bursts and flavor mixing during their travel to the Earth. However, since the information on the initial composition requires better precision in mixing angles and vice versa, neutrino detection at telescopes cannot provide solutions to both problems by itself. Thus, a probability to be measured at long-baseline oscillation is considered as a complement to the telescope, and problems like source identification and parameter degeneracy are examined under a few assumptions.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Energy-efficient, Large-scale Molecular Dynamics Simulations via Hardware- and Algorithm-level Optimization

This work aims to develop a framework for energy-efficient computing that will enable molecular dynamics (MD) simulations of large-scale phenomena with atomic precision and simultaneously remove computational bottlenecks limiting the speed of MD simulations. We seek to implement such an approach through the development of surrogate models for the interatomic force calculation combined with the use of mixed numerical precision formats. For a model system of neutral atoms (only pairwise interactions), significant force calculation efficiency improvements were achieved, without detrimental effects on atomic structures or average energies, using single precision, by developing a surrogate model (deep neural network), and by quantizing this surrogate model. For a model system of charged atoms, the reciprocal-space calculation of electrostatic interactions was identified as the main bottleneck, and the development of a surrogate model should be pursued to achieve an estimated one-order-of-magnitude additional speedup.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Compressed basis GMRES on high-performance graphics processing units

Krylov methods provide a fast and highly parallel numerical tool for the iterative solution of many large-scale sparse linear systems. To a large extent, the performance of practical realizations of these methods is constrained by the communication bandwidth in current computer architectures, motivating the investigation of sophisticated techniques to avoid, reduce, and/or hide the message-passing costs (in distributed platforms) and the memory accesses (in all architectures). This article leverages Ginkgo’s memory accessor in order to integrate a communication-reduction strategy into the (Krylov) GMRES solver that decouples the storage format (i.e., the data representation in memory) of the orthogonal basis from the arithmetic precision that is employed during the operations with that basis. Given that the execution time of the GMRES solver is largely determined by the memory accesses, the cost of the datatype transforms can be mostly hidden, resulting in the acceleration of the iterative step via a decrease in the volume of bits being retrieved from memory. Together with the special properties of the orthonormal basis (whose elements are all bounded by 1), this paves the road toward the aggressive customization of the storage format, which includes some floating-point as well as fixed-point formats with mild impact on the convergence of the iterative process. We develop a high-performance implementation of the “compressed basis GMRES” solver in the Ginkgo sparse linear algebra library using a large set of test problems from the SuiteSparse Matrix Collection. We demonstrate robustness and performance advantages on a modern NVIDIA V100 graphics processing unit (GPU) of up to 50% over the standard GMRES solver that stores all data in IEEE double-precision.

97 MATHEMATICS AND COMPUTING↗

Hybrid nanocapsules for in situ TEM imaging of gas evolution reactions in confined liquids

Liquid cell transmission electron microscopy (TEM) enables the direct observation of dynamic physical and chemical processes in liquids at the nanoscale. Quantitative investigations into reactions with fast kinetics and/or multiple reagents will benefit from further advances in liquid cell design that facilitate rapid in situ mixing and precise control over reagent volumes and concentrations. This work reports the development of inorganic–organic nanocapsules for high-resolution TEM imaging of nanoscale reactions in liquids with well-defined zeptoliter volumes. These hybrid nanocapsules, with 48 nm average diameter, consist of a thin layer of gold coating a lipid vesicle. As a model reaction, the nucleation, growth, and diffusion of nanobubbles generated by the radiolysis of water is investigated inside the nanocapsules. When the nanobubbles are sufficiently small (10–25 nm diameter), they are mobile in the nanocapsules, but their movement deviates from Brownian motion, which may result from geometric confinement by the nanocapsules. Gases and fluids can be transported between two nanocapsules when they fuse, demonstrating in situ mixing without using complex microfluidic schemes. The ability to synthesize nanocapsules with controlled sizes and to monitor dynamics simultaneously inside multiple nanocapsules provides opportunities to investigate nanoscale processes such as single nanoparticle synthesis in confined volumes and biological processes such as biomineralization and membrane dynamics.

36 MATERIALS SCIENCE↗

Analysis of Vector Particle-In-Cell (VPIC) memory usage optimizations on cutting-edge computer architectures

Vector Particle-In-Cell (VPIC) is one of the fastest plasma simulation codes in the world, with particle numbers ranging from one trillion on the first petascale system, Roadrunner, to ten trillion particles on the more recent Blue Waters supercomputer. As supercomputers continue to grow rapidly in size, so too does the gap between computing capability and memory capability. Current memory systems limit VPIC simulations greatly as the maximum number of particles that can be simulated directly depends on the available memory. In this study, we present a suite of VPIC memory optimizations (i.e., particle weight, half-precision, and fixed-point optimizations) that enable a significant increase in the number of particles in VPIC simulations. Here, we assess the optimizations’ impact on memory and runtime performance for a suite of cutting-edge computer architectures such has the NVIDIA V100 GPU, the IBM Power9, and the Fujitsu A64FX architectures. Our optimizations enable a 31.25% reduction in memory usage and up to 40% increase in the number of particles. This paper extends our work on developing particle storage format optimizations Tan et al.

97 MATHEMATICS AND COMPUTING↗

DFT-FE – A massively parallel adaptive finite-element code for large-scale density functional theory calculations

We present an accurate, efficient and massively parallel finite-element code, DFT-FE, for large-scale ab-initio calculations (reaching ~100,000 electrons) using Kohn-Sham density functional theory (DFT). DFT-FE is based on a local real-space variational formulation of the Kohn-Sham DFT energy functional that is discretized using a higher-order adaptive spectral finite-element (FE) basis, and treats pseudopotential and all-electron calculations in the same framework, while accommodating non-periodic, semi-periodic and periodic boundary conditions. We discuss the main aspects of the code, which include, the various strategies of adaptive FE basis generation, and the different approaches employed in the numerical implementation of the solution of the discrete Kohn-Sham problem that are focused on significantly reducing the floating point operations, communication costs and latency. We demonstrate the accuracy of DFT-FE by comparing the energies, ionic forces and periodic cell stresses on a wide range of problems with popularly used DFT codes. Further, we demonstrate that DFT-FE significantly outperforms widely used plane-wave codes---both in CPU-times and wall-times, and on both non-periodic and periodic systems---at systems sizes beyond a few thousand electrons, with over 5-10 fold speedups in systems with more than 10,000 electrons. In conclusion, the benchmark studies also highlight the excellent parallel scalability of DFT-FE, with strong scaling demonstrated on up to 192,000 MPI tasks.

97 MATHEMATICS AND COMPUTING↗

DFT-FE 1.0: A massively parallel hybrid CPU-GPU density functional theory code using finite-element discretization

In this work, we present DFT-FE 1.0, building on DFT-FE 0.6 [Comput. Phys. Commun. 246, 106853 (2020)], to conduct fast and accurate large-scale density functional theory (DFT) calculations (reaching ~ 100,000 electrons) on both many-core CPU and hybrid CPU-GPU computing architectures. This work involves improvements in the real-space formulation—via an improved treatment of the electrostatic interactions that substantially enhances the computational efficiency—as well high-performance computing aspects, including the GPU acceleration of all the key compute kernels in DFT-FE. We demonstrate the accuracy by comparing the ground-state energies, ionic forces and cell stresses on a wide-range of benchmark systems against those obtained from widely used DFT codes. Further, we demonstrate the numerical efficiency of our implementation, which yields ~ 20× CPU-GPU speed-up by using GPU acceleration on hybrid CPU-GPU nodes. Notably, owing to the parallel-scaling of the GPU implementation, we obtain wall-times of 80–140 seconds for full ground-state calculations, with stringent accuracy, on benchmark systems containing ~ 6, 000 – 15,000 electrons.

pseudopotential↗

Analysis of Femtosecond Timing Noise and Stability in Microwave Components

To probe chemical dynamics, X-ray pump-probe experiments trigger a change in a sample with an optical laser pulse, followed by an X-ray probe. At the Linac Coherent Light Source, LCLS, timing differences between the optical pulse and x-ray probe have been observed with an accuracy as low as 50 femtoseconds. This sets a lower bound on the number of frames one can arrange over a time scale to recreate a 'movie' of the chemical reaction. The timing system is based on phase measurements from signals corresponding to the two laser pulses; these measurements are done by using a double-balanced mixer for detection. To increase the accuracy of the system, this paper studies parameters affecting phase detection systems based on mixers, such as signal input power, noise levels, temperature drift, and the effect these parameters have on components such as the mixers, splitters, amplifiers, and phase shifters. Noise data taken with a spectrum analyzer show that splitters based on ferrite cores perform with less noise than strip-line splitters. The data also shows that noise in specific mixers does not correspond with the changes in sensitivity per input power level. Temperature drift is seen to exist on a scale between 1 and 27 fs/{sup o}C for all of the components tested. Results show that any components using more metallic conductor tend to exhibit more noise as well as more temperature drift. The scale of these effects is large enough that specific care should be given when choosing components and designing the housing of high precision microwave mixing systems for use in detection systems such as the LCLS. With these improvements, the timing accuracy can be improved to lower than currently possible.

43 PARTICLE ACCELERATORS↗