Search NASA⌕ Search

SEARCH · Search NASA

Results for “core 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 1,801 records · Page 100

5.1.6 Demo #6: NanoStitch (NAWAStitch) Multifunctional Composites Trade Validation

The overall objective of the NanoStitch Multifunctional Composites Trade Validation Demonstration project was to determine the value proposition of adding NanoStitch to rapid-cure composite systems for structures used in the expanding sector of advanced air mobility (AAM) and specifically electric vertical take-off and landing (eVTOL) aircraft. The expected value proposition was stronger composite parts, with new multifunctional attributes, that are lower cost to fabricate and maintain. NanoStitch is a new composite innovation where a contiguous vertically aligned carbon nanotube (VACNT) forest is located between each ply of a laminate composite resulting in improved mechanical properties as well as other multifunctional properties such as enhanced thermal and electrical conductivity. NanoStitch was originally distributed by N12, later acquired by NAWA Technologies, now called NAWAStitch, and henceforth referred to in this document as NAWAStitch. Two significant evaluations were framed to help meet the overall objective. First, was Task 1 to evaluate the effects that adding NAWAStitch to a rapid cure material system would have on the mechanical and processing properties of the material. Second, was Task 2 to evaluate the workability of adding NAWAStitch to practical flight demonstration articles. From previous research, it was already known that the application of NAWAStitch to a composite structure can enhance the mechanical properties of that structure. However, this project will help the engineering community understand those benefits and the basic structural strength values to apply when incorporating NAWAStitch material into composite structure that utilizes a rapid-cure resin system, which is novel. Testing showed that by adding NAWAStitch to the prepreg, most material properties were improved because the VACNTs strengthened the resin to fiber interface, and the interface itself, between plies. For example compression strength improved by more than 20%, and erosion resistance by more than 50%. The UDRI team identified two flight demonstration articles through discussions with Aurora Flight Sciences (Aurora) and reviewing literature on air mobility and the eVTOL market. The demonstration articles were a thin skin honeycomb panel and a section from an eVTOL propeller blade. The manufacturing and testing trials of the demonstration articles were as follows: • NAWAStitch added no manufacturing complexity or detriment to the article physical qualities, while at the same time added multifunctional properties such as enhanced rain erosion and electrothermal de-icing besides enhanced mechanical properties • VACNTs did not increase or hurt the mechanical properties of the interface between honeycomb core materials and the thin skin face sheet • Demonstrated successful de-icing function in an eVTOL propeller meeting real-world specifications in uniformity and endurance • Multifunctional eVTOL propeller cost analysis showed a cost savings of at least 67 % for a single propeller blade with multifunctional de-icing capability and improved mechanical properties which equates to a potential savings of $10,116 per eVTOL aircraft with six three-bladed propellers Overall, the results of this project can provide additional tools for the composite industry to optimize energy efficiency through faster and stronger composite parts that are lower cost to fabricate and maintain. Additionally, the results of this effort can help US industry like NAWA America set the stage for rapid growth and establishment of a manufacturing facility in Ohio. The project has given NAWA America the basic understanding of how the addition of VACNTs to a composite structure can affect the mechanical properties and hence point them to potential commercial applications. The project has also provided Teijin Carbon US with another potential material application for their rapid cure materials that are manufactured at the Renegade facility in Miamisburg, OH.

36 MATERIALS SCIENCE↗

Precise Tuning of Surface Quenching for Luminescence Enhancement in Core–Shell Lanthanide-Doped Nanocrystals

Lanthanide-doped nanocrystals are of particular interest for the research community not only due to their ability to shape light by downshifting, quantum cutting, and upconversion but also because novel optical properties can be found by the precise engineering of core-shell nanocrystals. Because of the large surface area-to-volume ratio of nanocrystals, the luminescence is typically suppressed by surface quenching. Here, we demonstrate a mechanism that exploits surface quenching processes to improve the luminescence of our core-shell lanthanide-doped nanocrystals. By carefully tuning the shell thickness of inert β-NaLuF 4 around β-NaYF 4 nanocrystals doped with Yb 3+ and Er 3+ , we unravel the relationship between quantum yield and shell thickness, and quantify surface quenching rates for the relevant Er 3+ and Yb 3+ energy levels. This enhanced understanding of the system's dynamics allowed us to design nanocrystals with a surface quenching-assisted mechanism for bright NIR to NIR downshifting with a distinctive efficiency peak for an optimized shell thickness.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Modeling and sensitivity analysis of mass transfer in active multilayer polymeric film for food applications

The barrier performance of multilayer polymeric films for food applications has been significantly improved by incorporating oxygen scavenging materials. The scavenging activity depends on parameters such as diffusion coefficient, solubility, concentration of scavenger loaded and the number of available reactive sites. These parameters influence the barrier performance of the film in different ways. Virtualization of the process is useful to characterize, design and optimize the barrier performance based on physical configuration of the films. Also, the knowledge of values of parameters is important to predict the performances. Inverse modeling and sensitivity analysis are sole way to find reasonable values of poorly defined, unmeasured parameters and to analyze the most influencing parameters. Thus, the objective of this work was to develop a model to predict barrier properties of multilayer film incorporated with reactive layers and to analyze and characterize their performances. Polymeric film based on three layers of Polyethylene terephthalate (PET), with a core reactive layer, at different thickness configurations was considered in the model. A one dimensional diffusion equation with reaction was solved numerically to predict the concentration of oxygen diffused into the polymer taking into account the reactive ability of the core layer. The model was solved using commercial software for different film layer configurations and sensitivity analysis based on inverse modeling was carried out to understand the effect of physical parameters. The results have shown that the use of sensitivity analysis can provide physical understanding of the parameters which highly affect the gas permeation into the film. Solubility and the number of available reactive sites were the factors mainly influencing the barrier performance of three layered polymeric film. Multilayer films slightly modified the steady transport properties in comparison to net PET, giving a small reduction in the permeability and oxygen transfer rate values. Scavenging capacity of the multilayer film increased linearly with the increase of the reactive layer thickness and the oxygen absorption reaction at short times decreased proportionally with the thickness of the external PET layer.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

First neutral beam experiments on Wendelstein 7-X

In the previous divertor campaign, the Wendelstein 7-X (W7-X) device injected 3.6 MW of neutral beam heating power allowing for the achievement of densities approaching 2 × 10 20 m -3 , and providing the first initial assessment of fast ion confinement in a drift optimized stellarator. The neutral beam injection (NBI) system on W7-X is comprised of two beam boxes with space for four radio frequency sources each. The 3.6 MW of heating reported in this work was achieved with two sources in the NI21 beam box. The effect of combined electron-cyclotron resonance heating (ECRH) and NBI was explored through a series of discharges varying both NBI and ECRH power. Discharges without ECRH saw a linear increase in the line-integrated plasma density, and strong peaking of the core density, over the discharge duration. The presence of 1 MW of ECRH power was found to be sufficient to control a continuous density rise during NBI operation. Simulations of fast ion wall loads were found to be consistent with experimental infrared camera images during operation. In general, NBI discharges were free from the presence of fast ion induced Alfvénic activity, consistent with low beam betas. These experiments provide data for future scenario development and initial assessment of fast-ion confinement in W7-X, a key topic of the project.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Designing Algorithms for the EMU Migrating-threads-based Architecture

The decades-old memory bottleneck problem for data-intensive applications is getting worse as the processor core counts continue to increase. Workloads with sparse memory access characteristics only achieve a fraction of a system’s total memory bandwidth. EMU architecture provides a radical approach to the issue by migrating the computational threads to the location where the data resides. The system enables access to a large PGAS-type memory for hundreds of nodes via a Cilk- based multi-threaded execution scheme.EMU architecture brings brand new challenges in application design and development. Data distribution and thread creation strategies play a crucial role in achieving optimal performance in the EMU platform. In this work, we identify several design considerations that need to be taken care of while developing applications for the new architecture and we evaluate their performance effects on the EMU-chick hardware. We also present a modified BFS algorithm for the EMU system and give experi- mental results for its execution on the platform.

Belviranli, Mehmet E.↗

Optimal SCR Control Using Data-Driven Models

We present an optimal control solution for the urea injection for a heavy-duty diesel (HDD) selective catalytic reduction (SCR). The approach taken here is useful beyond SCR and could be applied to any system where a control strategy is desired and input-output data is available. For example, the strategy could also be used for the diesel oxidation catalyst (DOC) system. In this paper, we identify and validate a one-step ahead Kalman state-space estimator for downstream NOx using the bench reactor data of an SCR core sample. The test data was acquired using a 2010 Cummins 6.7L ISB production engine with a 2010 Cummins production aftertreatment system. We used a surrogate HDD federal test procedure (FTP), developed at Michigan Technological University (MTU), which simulates the representative transients of the standard FTP cycle, but has less engine speed/load points. The identified state-space model is then used to develop a tunable cost function that simultaneously minimizes NOx emissions and urea usage. The cost function is quadratic and univariate, thus the minimum can be computed analytically. We show the performance of the closed-loop controller in using a reduced-order discrete SCR simulator developed at MTU. Our experiments with the surrogate HDD-FTP data show that the strategy developed in this paper can be used to identify performance bounds for urea dose controllers.

diesel, emmissions, control, environment, modeling↗

A diamond detector for inertial confinement fusion X-ray bang-time measurements at the National Ignition Facility

An instrument has been developed to measure X-ray bang-time for inertial confinement fusion capsules; the time interval between the start of the laser pulse and peak X-ray emission from the fuel core. The instrument comprises chemical vapor deposited polycrystalline diamond photoconductive X-ray detectors with highly ordered pyrolytic graphite X-ray monochromator crystals at the input. Capsule bang-time can be measured in the presence of relatively high thermal and hard X-ray background components due to the selective band pass of the crystals combined with direct and indirect X-ray shielding of the detector elements. A five channel system is being commissioned at the National Ignition Facility at Lawrence Livermore National Laboratory for implosion optimization measurements as part of the National Ignition Campaign. Characteristics of the instrument have been measured demonstrating that X-ray bang-time can be measured with {+-} 30ps precision, characterizing the soft X-ray drive to +/- 1eV or 1.5%.

42 ENGINEERING↗

Progress towards accelerating HOMME on hybrid multi- core systems

The suitability of a spectral element based dynamical core (HOMME) within the Community Atmospheric Model (CAM) for GPU-based architectures is examined and initial performance results are reported. This work was done within a project to enable CAM to run at high resolution on next generation, multi-petaflop systems. The dynamical core is the present focus because it dominates the performance profile of our target problem. HOMME enjoys good scalability due to its underlying cubed-sphere mesh with full two-dimensional decomposition and the localization of all computational work within each element. The thread blocking and code changes that allow HOMME to effectively use GPUs are described along with a rewritten vertical remapping scheme which improves performance on both CPUs and GPUs. Validation of results in the full HOMME model is also described. Remaining issues affecting performance include optimizing the boundary exchanges for the case of multiple spectral elements being computed on the GPU and using multiple CUDA streams to overlap data transfers with computations.

99 GENERAL AND MISCELLANEOUS↗

Enabling Dynamic Probabilistic Risk Assessment of Physical Security Using EMRALD and MAAP (Presentation)

The optimization of physical security in nuclear power plants requires sophisticated methodologies that integrate operator actions and plant behavior through advanced simulation tools. Idaho National Laboratory has developed the Modeling and Analysis for Safety and Security using the Dynamic EMRALD Framework (MASS-DEF) methodology, an approach that integrates force-on-force simulations, dynamic probabilistic risk assessment, and thermal hydraulics modeling to enhance security planning while reducing costs. A reduced order model for thermal hydraulic simulations performed by the Modular Accident Analysis Program (MAAP) was developed to evaluate reactor core behavior during attack scenarios. MAAP simulations are computationally intensive and must be run in a secure environment, complicating analysis and validation. By pre-computed scenario outcomes for a small number of modified parameters, the reduced order model significantly decreases the computational cost and enables offsite review of the results.

22 - GENERAL STUDIES OF NUCLEAR REACTORS↗

Runtime power allocation approach for GAMESS hybrid CPU‐GPU implementation

Summary To improve power consumption of applications at the runtime, modern processors provide frequency scaling capabilities, which along with workload optimization, are also available on GPU accelerators. In this work, a runtime strategy is proposed to distribute a given power allocation among the host components and the GPU according to the current application performance and power usage, such that GPU execution is prioritized over CPU for power allocation to maximize application performance. Next, the strategy is tailored to an application, a quantum‐chemistry package GAMESS for ab initio electronic structure calculations. Specifically, GAMESS hybrid CPU–GPU implementation as provided in the Libcchem library is considered. Experiments, performed on a 28‐core node with a Kepler GPU, resulted in performance gains of up to 50% under the proposed strategy and the largest power allocation considered here as compared with the scenario when this allocation was equally distributed among the computing‐platform components.

Sundriyal, Vaibhav↗

An important criterion for reliable multi-monochromatic x-ray imager diagnostics and its impact on the reconstructed images

Temperature and density asymmetry diagnosis is critical to advance inertial confinement fusion (ICF) science. A multimonochromatic x-ray imager, MMI, records the spectral signature from an ICF implosion core with time resolution, 2D spatial resolution and spectral resolution. While narrow-band images and 2D space-resolved spectra from the MMI data constrain the temperature and the density spatial structure of the core, the accuracy of the images and the spectra highly depends on the quality of the MMI data and the processing tools. Here, we synthetically investigate the criterion for reliable MMI diagnostics and its effects on the accuracy of the reconstructed images. The pinhole array tilt determines the object spatial sampling efficiency and the minimum reconstruction width, $\textit{w}$. When the spectral width associated with $\textit{w}$ is significantly narrower than the spectral linewidth, the line images reconstructed from the MMI data become reliable. The MMI setup has to be optimized for every application to meet this criterion for reliable ICF diagnostics.

47 OTHER INSTRUMENTATION↗

Inclusion of trial functions in the Langevin equation path integral ground state method: Application to parahydrogen clusters and their isotopologues

We developed and studied the implementation of trial wavefunctions in the newly proposed Langevin equation Path Integral Ground State (LePIGS) method [S. Constable, M. Schmidt, C. Ing, T. Zeng, and P.-N. Roy, J. Phys. Chem. A 117, 7461 (2013)]. The LePIGS method is based on the Path Integral Ground State (PIGS) formalism combined with Path Integral Molecular Dynamics sampling using a Langevin equation based sampling of the canonical distribution. This LePIGS method originally incorporated a trivial trial wavefunction, ψ{sub T}, equal to unity. The present paper assesses the effectiveness of three different trial wavefunctions on three isotopes of hydrogen for cluster sizes N = 4, 8, and 13. The trial wavefunctions of interest are the unity trial wavefunction used in the original LePIGS work, a Jastrow trial wavefunction that includes correlations due to hard-core repulsions, and a normal mode trial wavefunction that includes information on the equilibrium geometry. Based on this analysis, we opt for the Jastrow wavefunction to calculate energetic and structural properties for parahydrogen, orthodeuterium, and paratritium clusters of size N = 4 − 19, 33. Energetic and structural properties are obtained and compared to earlier work based on Monte Carlo PIGS simulations to study the accuracy of the proposed approach. The new results for paratritium clusters will serve as benchmark for future studies. This paper provides a detailed, yet general method for optimizing the necessary parameters required for the study of the ground state of a large variety of systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

FINDING {eta} CAR ANALOGS IN NEARBY GALAXIES USING SPITZER. I. CANDIDATE SELECTION

The late-stage evolution of the most massive stars such as {eta} Carinae is controlled by the effects of mass loss, which may be dominated by poorly understood eruptive mass ejections. Understanding this population is challenging because no true analogs of {eta} Car have been clearly identified in the Milky Way or other galaxies. We utilize Spitzer IRAC images of seven nearby ({approx}< 4 Mpc) galaxies to search for such analogs. We find 34 candidates with a flat or rising mid-IR spectral energy distributions toward longer mid-infrared wavelengths that emit >10{sup 5} L{sub Sun} in the IRAC bands (3.6 to 8.0 {mu}m) and are not known to be background sources. Based on our estimates for the expected number of background sources, we expect that follow-up observations will show that most of these candidates are not dust enshrouded massive stars, with an expectation of only 6 {+-} 6 surviving candidates. Since we would detect true analogs of {eta} Car for roughly 200 years post-eruption, this implies that the rate of eruptions like {eta} Car is less than the core-collapse supernova rate. It is possible, however, that every M > 40 M{sub Sun} star undergoes such eruptions given our initial results. In Paper II we will characterize the candidates through further analysis and follow-up observations, and there is no barrier to increasing the galaxy sample by an order of magnitude. The primary limitation of the present search is that Spitzer's resolution limits us to the shorter wavelength IRAC bands. With the James Webb Space Telescope, such surveys can be carried out at the far more optimal wavelengths of 10-30 {mu}m, allowing identification of {eta} Car analogs for millennia rather than centuries post-eruption.

79 ASTRONOMY AND ASTROPHYSICS↗

GPU-Based Static Data-Flow Analysis for Fast and Scalable Android App Vetting

Many popular vetting tools for Android applications use static code analysis techniques. In particular, Interprocedural Data-Flow Graph (IDFG) construction is the computation at the core of Android static data-flow analysis and consumes most of the analysis time. Many analysis tools use a worklist algorithm, an iterative fixed-point approach, to construct the IDFG. In this paper, we observe that a straightforward GPU parallelization of the worklist algorithm leads to significant underutilization of the GPU resources. We identify four performance bottlenecks, namely, frequent dynamic memory allocations, high branch divergence, workload imbalance, and irregular memory access patterns. Accordingly, we propose GDroid, a GPU-based worklist algorithm implementation with multiple fine-grained optimizations tailored to common characteristics of Android applications. The optimizations considered are: matrix-based data structure, memory access-based node grouping, and worklist merging. Our experimental evaluation, performed on 1000 Android applications, shows that the proposed optimizations are beneficial to performance, and GDroid can achieve up to 128X speedups against a plain GPU implementation.

Android security↗

SU-F-BRD-02: Application of ARCHERRT-- A GPU-Based Monte Carlo Dose Engine for Radiation Therapy -- to Tomotherapy and Patient-Independent IMRT

Purpose: As a module of ARCHER -- Accelerated Radiation-transport Computations in Heterogeneous EnviRonments, ARCHER{sub RT} is designed for RadioTherapy (RT) dose calculation. This paper describes the application of ARCHERRT on patient-dependent TomoTherapy and patient-independent IMRT. It also conducts a 'fair' comparison of different GPUs and multicore CPU. Methods: The source input used for patient-dependent TomoTherapy is phase space file (PSF) generated from optimized plan. For patient-independent IMRT, the open filed PSF is used for different cases. The intensity modulation is simulated by fluence map. The GEANT4 code is used as benchmark. DVH and gamma index test are employed to evaluate the accuracy of ARCHER{sub RT} code. Some previous studies reported misleading speedups by comparing GPU code with serial CPU code. To perform a fairer comparison, we write multi-thread code with OpenMP to fully exploit computing potential of CPU. The hardware involved in this study are a 6-core Intel E5-2620 CPU and 6 NVIDIA M2090 GPUs, a K20 GPU and a K40 GPU. Results: Dosimetric results from ARCHER{sub RT} and GEANT4 show good agreement. The 2%/2mm gamma test pass rates for different clinical cases are 97.2% to 99.7%. A single M2090 GPU needs 50~79 seconds for the simulation to achieve a statistical error of 1% in the PTV. The K40 card is about 1.7∼1.8 times faster than M2090 card. Using 6 M2090 card, the simulation can be finished in about 10 seconds. For comparison, Intel E5-2620 needs 507∼879 seconds for the same simulation. Conclusion: We successfully applied ARCHER{sub RT} to Tomotherapy and patient-independent IMRT, and conducted a fair comparison between GPU and CPU performance. The ARCHER{sub RT} code is both accurate and efficient and may be used towards clinical applications.

60 APPLIED LIFE SCIENCES↗

Control And Optimization Modular Modeling Application For Nuclear Deployment

The purpose of the COMMAND code is to provide a flexible, scalable tool for use in developing, integrating, and testing the technologies necessary for achieving autonomous operations of advanced nuclear reactors. The code enables users to efficiently implement custom simulations and experiments by combining key methods from different software modules. These modules are focused on: modeling and simulation tools, such as nuclear simulation tools used for high-fidelity modeling (e.g., Reactor Excursion and Leak Analysis Program [RELAP5-3D] and Monte Carlo N-Particle [MCNP]); machine learning and optimization tools (e.g., anomaly detection and data-driven modeling techniques); advanced control in its digital, high-performance, and supervisory control forms (e.g., proportional integral derivative (PID) control and model predictive control (MPC); and integration with hardware through industrial communication protocols. To ensure flexibility and scalability, COMMAND was designed to be both modular—the software “pieces” all inherit from generic building blocks and can be combined and connected to create complicated simulations—and high performing—designed for parallel processing, enabling simulations and experiments to take advantage of multi-core computers, servers, and nodes. The code is written in the Python programming language due to the language's popularity, active community, and open-source and cross-platform nature. Maintaining consistency with other simulation tools used within the nuclear energy community, users implement simulations and experiments through text input files, which define components, parameters, connections, etc., through lines of text. Given that COMMAND is written in Python, these input files are native Python scripts, and so use the standard Python structure and formatting. This also enables users to take advantage of Python's extensive package library to develop custom capabilities for their specific use cases.

Faber, Jacob [Idaho National Laboratory (INL), Ida↗

BayFlux: A Bayesian method to quantify metabolic Fluxes and their uncertainty at the genome scale

Metabolic fluxes, the number of metabolites traversing each biochemical reaction in a cell per unit time, are crucial for assessing and understanding cell function. 13 C Metabolic Flux Analysis ( 13 C MFA) is considered to be the gold standard for measuring metabolic fluxes. 13 C MFA typically works by leveraging extracellular exchange fluxes as well as data from 13 C labeling experiments to calculate the flux profile which best fit the data for a small, central carbon, metabolic model. However, the nonlinear nature of the 13 C MFA fitting procedure means that several flux profiles fit the experimental data within the experimental error, and traditional optimization methods offer only a partial or skewed picture, especially in “non-gaussian” situations where multiple very distinct flux regions fit the data equally well. Here, we present a method for flux space sampling through Bayesian inference (BayFlux), that identifies the full distribution of fluxes compatible with experimental data for a comprehensive genome-scale model. This Bayesian approach allows us to accurately quantify uncertainty in calculated fluxes. We also find that, surprisingly, the genome-scale model of metabolism produces narrower flux distributions (reduced uncertainty) than the small core metabolic models traditionally used in 13 C MFA. The different results for some reactions when using genome-scale models vs core metabolic models advise caution in assuming strong inferences from 13 C MFA since the results may depend significantly on the completeness of the model used. Based on BayFlux, we developed and evaluated novel methods (P- 13 C MOMA and P- 13 C ROOM) to predict the biological results of a gene knockout, that improve on the traditional MOMA and ROOM methods by quantifying prediction uncertainty.

59 BASIC BIOLOGICAL SCIENCES↗

Revisiting Huffman Coding: Toward Extreme Performance on Modern GPU Architectures

Today's high-performance computing (HPC) applications are producing vast volumes of data, which are challenging to store and transfer efficiently during the execution, such that data compression is becoming a critical technique to mitigate the storage burden and data movement cost. Huffman coding is arguably the most efficient Entropy coding algorithm in information theory, such that it could be found as a fundamental step in many modern compression algorithms such as DEFLATE. On the other hand, today's HPC applications are more and more relying on the accelerators such as GPU on supercomputers, while Huffman encoding suffers from low throughput on GPUs, resulting in a significant bottleneck in the entire data processing. In this paper, we propose and implement an efficient Huffman encoding approach based on modern GPU architectures, which addresses two key challenges: (1) how to parallelize the entire Huffman encoding algorithm, including codebook construction, and (2) how to fully utilize the high memory-bandwidth feature of modern GPU architectures. The detailed contribution is fourfold. (1) We develop an efficient parallel codebook construction on GPUs that scales effectively with the number of input symbols. (2) We propose a novel reduction based encoding scheme that can efficiently merge the codewords on GPUs. (3) We optimize the overall GPU performance by leveraging the state-of-the-art CUDA APIs such as Cooperative Groups. (4) We evaluate our Huffman encoder thoroughly using six real-world application datasets on two advanced GPUs and compare with our implemented multithreaded Huffman encoder. Experiments show that our solution can improve the encoding throughput by up to 5.0× and 6.8× on NVIDIA RTX 5000 and V100, respectively, over the state-of-the-art GPU Huffman encoder, and by up to 3.3× over the multithread encoder on two 28-core Xeon Platinum 8280 CPUs.

Tian, Jiannan↗

Refine your search to explore more results.