Results for “Data transfer node” — Page 7

NASA NTRSDate not supplied

Numerical Modeling of No Vent Filling of a Cryogenic Tank with Thermo-dynamic Vent System Assisted Injector

This paper presents a multi-node finite volume model of No Vent Filling (NVF) of a cryogenic tank with Thermo-dynamic Vent System (TVS) assisted injector, using the Generalized Fluid System Simulation Program (GFSSP), a general purpose flow network code. NVF tests were conducted in a CRYOgenic Orbital Testbed (CRYOTE) tank which was filled by liquid nitrogen using a TVS assisted injector with vent valve closed during the entire filling process. In TVS assisted injector, the liquid flow splits into two streams: one stream is routed to a Joule-Thomson (J-T) orifice where the flow immediately flashes from liquid to vapor or two phase mixture because the downstream to J-T leg is maintained at vacuum level.; the other stream is injected into the tank after being cooled by cold vapor of J-T leg in a heat exchanger. The flow through J-T leg is also used to cool the outer metal matrix of the injector which in turn cools the vapor in the tank ullage. The cooling of vapor in the ullage by cold injector surface reduces ullage pressure that allows liquid to enter and fill the tank. An integrated numerical model of the test set up was developed. The model included the dual lines of TVS assisted injector, tank and tank wall. The tank was discretized into multiple fluid nodes and branches to represent the ullage and liquid nitrogen and multiple solid nodes to represent the tank wall and structure. The heat transfer between solid to fluid was calculated from pool boiling correlations which include film, transition, and nucleate boiling, as well as natural convection during pre- and post-boiling. The model also accounts for the condensation of vapor at ullage-injector interface and when it comes in contact with the liquid spray. The predicted pressure, resident mass, wall temperature in the tank were compared with the test data.

Alok Majumdar↗

DOE OSTI2020

Opportunities for Cost Savings with In-Transit Visualization

We analyze the opportunities for in-transit visualization to provide cost savings compared to in-line visualization. We begin by developing a cost model that includes factors related to both in-line and in-transit which allows comparisons to be made between the two methods. We then run a series of studies to create a corpus of data for our model. We run two different visualization algorithms, one that is computation heavy and one that is communication heavy with concurrencies up to 32, 768 cores. Our primary results are in exploring the cost model within the context of our corpus. Our findings show that in-transit consistently achieves significant cost efficiencies by running visualization algorithms at lower concurrency, and that in many cases these efficiencies are enough to offset other costs (transfer, blocking, and additional nodes) to be cost effective overall. Finally, this work informs future studies, which can focus on choosing ideal configurations for in-transit processing that can consistently achieve cost efficiencies.

Kress, James↗

DOE OSTI2025

Accelerating high-order continuum kinetic plasma simulations using multiple GPUs

Kinetic plasma simulations solve the Vlasov-Poisson or Vlasov-Maxwell equations to evolve scalar-variable distribution functions in position-velocity phase space and vector-variable electromagnetic fields in configuration space. The immense computational cost of evolving high-dimensional variables, and their large number of degrees of freedom, often limits the utility of continuum kinetic simulations and presents a challenge when it comes to accurately simulating real-world physical phenomena. To address this challenge, we present techniques that accelerate and minimize the computational work required for a scalable Vlasov-Poisson solver. We show theoretical hardware compute and communication bounds for solving a fourth-order finite-volume Vlasov-Poisson system. These bounds are then used to inform and evaluate the design of performance portable algorithms for a multiple graphics processing unit (GPU) accelerated version of the Vlasov-Poisson solver VCK-CPU [1]. We demonstrate that the multi-GPU Vlasov solver implementation, VCK-GPU, simultaneously minimizes required inter-process data transfer while also being bounded by the machine network performance limits. This results in an overall strong scaling speedup per timestep of up to 40x in three-dimensional phase space (one position, two velocity coordinates) and 54x in four dimensional phase space (two position, two velocity coordinates) and a 341x increase in simulation throughput of the GPU accelerated code over the existing CPU code. The GPU code is also able to weak scale up to 256 compute nodes and 1024 GPUs. In conclusion, we demonstrate that the improved compute performance enables exploring configurations which were previously computationally infeasible, including resolving fine-scale distribution function filamentation and multi-species dynamics with realistic electron-proton mass ratios.

Continuum kinetics↗

DOE OSTI2021

Enabling Scalable and Extensible Memory-mapped Datastores in Userspace

Exascale workloads are expected to incorporate data-intensive processing in close coordination with traditional physics simulations. These emerging scientific, data-analytics and machine learning applications need to access a wide variety of datastores in flat files and structured databases. Programmer productivity is greatly enhanced by mapping datastores into the application process's virtual memory space to provide a unified “in-memory” interface. Currently, memory mapping is provided by system software primarily designed for generality and reliability. However, scalability at high concurrency is a formidable challenge on exascale systems. Also, there is a need for extensibility to support new datastores potentially requiring HPC data transfer services. In this article, we present UMap , a scalable and extensible userspace service for memory-mapping datastores. Furthermore, through decoupled queue management, concurrency aware adaptation, and dynamic load balancing, UMap enables application performance to scale even at high concurrency. We evaluate UMap in data-intensive applications, including sorting, graph traversal, database operations, and metagenomic analytics. Our results show that UMap as a userspace service outperforms an optimized kernel-based service across a wide range of intra-node concurrency by 1.22-1.9 × . We performed two case studies to demonstrate UMap 's extensibility. First, a new datastore residing in remote memory is incorporated into UMap as an application-specific plugin. Second, we present a persistent memory allocator Metall built atop UMap for unified storage/memory.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS2024

Nodal Modeling of Liquid Propellant Feed and Pressurization System

Nodal or network flow modeling plays an important role in the design and operation of the feed and pressurization system of a liquid rocket engine. Model development and execution time is relatively short for nodal codes in comparison to the Navier-Stokes based CFD codes. Nodal models also allow the inclusion of several components into one model to predict the behavior of a larger system. Unlike CFD models, the nodal models do not need very fine discretization of the flow field because they use empirical correlations to model fluid friction and heat transfer. This paper presents several applications of nodal modeling of liquid propellant feed and pressurization systems using the Generalized Fluid System Simulation Program (GFSSP), a nodal code developed at NASA/ Marshall Space Flight Center. GFSSP discretizes the flow field into nodes which are connected by branches. The mass and energy conservation equations and the equation of state are solved to calculate pressure, temperature, and resident mass at the nodes. The momentum equations are solved at the branches to calculate flow rates. Applications include a) tank pressurization by inert gas as well as autogenous pressurization by gaseous propellant, b) submerged pressurization by helium, c) self-pressurization due to boil-off of cryogenic propellant, d) chilldown of a transfer line of a cryogenic propellant feed system, and e) chilldown and filling of a cryogenic tank. Each of the above-mentioned models was verified and validated by comparing with test data.

Nodal Model↗

NASA NTRS2024

Nodal Modeling of Liquid Propellant Feed and Pressurization System

Nodal or network flow modeling plays an important role in the design and operation of the feed and pressurization system of a liquid rocket engine. Model development and execution time is relatively short for nodal codes in comparison to the Navier-Stokes based CFD codes. Nodal models also allow the inclusion of several components into one model to predict the behavior of a larger system. Unlike CFD models, the nodal models do not need very fine discretization of the flow field because they use empirical correlations to model fluid friction and heat transfer. This paper presents several applications of nodal modeling of liquid propellant feed and pressurization systems using the Generalized Fluid System Simulation Program (GFSSP), a nodal code developed at NASA/ Marshall Space Flight Center. GFSSP discretizes the flow field into nodes which are connected by branches. The mass and energy conservation equations and the equation of state are solved to calculate pressure, temperature, and resident mass at the nodes. The momentum equations are solved at the branches to calculate flow rates. Applications include a) tank pressurization by inert gas as well as autogenous pressurization by gaseous propellant, b) submerged pressurization by helium, c) self-pressurization due to boil-off of cryogenic propellant, d) chilldown of a transfer line of a cryogenic propellant feed system, and e) chilldown and filling of a cryogenic tank. Each of the above-mentioned models was verified and validated by comparing with test data.

Nodal Model↗

DOE OSTI2021

Coding the Computing Continuum: Fluid Function Execution in Heterogeneous Computing Environments

Advances in network technologies have greatly decreased barriers to accessing physically distributed computers. This newfound accessibility coincides with increasing hardware specialization, creating exciting new opportunities to dispatch workloads to the best resource for a specific purpose, rather than those that are closest or most easily accessible. We present Delta, a service designed to intelligently schedule function-based workloads across a distributed set of heterogeneous computing resources. Delta implements an extensible architecture in which different predictors and scheduling algorithms can be integrated to provide dynamically evolving estimates of function execution times on different resources-estimates that can be used to determine the most appropriate location for execution. We describe predictors for function runtime, data transfer time, and cold-start resource provisioning and configuration delay; dynamic learning methods that update predictor models over time; and scheduling strategies that take into account both function and endpoint information. We show that these methods can halve workload makespan when compared with a strategy that selects the fastest resource, and decrease makespan by a factor of five when compared to a round robin strategy, when deployed on a heterogeneous testbed with resources ranging from a Raspberry Pi to a GPU node in an academic cloud.

Computing continuum↗

DOE OSTI2023

Porting the WAVEWATCH III (v6.07) wave action source terms to GPU

Abstract. Surface gravity waves play a critical role in several processes, including mixing, coastal inundation, and surface fluxes. Despite the growing literature on the importance of ocean surface waves, wind–wave processes have traditionally been excluded from Earth system models (ESMs) due to the high computational costs of running spectral wave models. The development of the Next Generation Ocean Model for the DOE’s (Department of Energy) E3SM (Energy Exascale Earth System Model) Project partly focuses on the inclusion of a wave model, WAVEWATCH III (WW3), into E3SM. WW3, which was originally developed for operational wave forecasting, needs to be computationally less expensive before it can be integrated into ESMs. To accomplish this, we take advantage of heterogeneous architectures at DOE leadership computing facilities and the increasing computing power of general-purpose graphics processing units (GPUs). This paper identifies the wave action source terms, W3SRCEMD, as the most computationally intensive module in WW3 and then accelerates them via GPU. Our experiments on two computing platforms, Kodiak (P100 GPU and Intel(R) Xeon(R) central processing unit, CPU, E5-2695 v4) and Summit (V100 GPU and IBM POWER9 CPU) show respective average speedups of 2× and 4× when mapping one Message Passing Interface (MPI) per GPU. An average speedup of 1.4× was achieved using all 42 CPU cores and 6 GPUs on a Summit node (with 7 MPI ranks per GPU). However, the GPU speedup over the 42 CPU cores remains relatively unchanged (∼ 1.3×) even when using 4 MPI ranks per GPU (24 ranks in total) and 3 MPI ranks per GPU (18 ranks in total). This corresponds to a 35 %–40 % decrease in both simulation time and usage of resources. Due to too many local scalars and arrays in the W3SRCEMD subroutine and the huge WW3 memory requirement, GPU performance is currently limited by the data transfer bandwidth between the CPU and the GPU. Ideally, OpenACC routine directives could be used to further improve performance. However, W3SRCEMD would require significant code refactoring to make this possible. We also discuss how the trade-off between the occupancy, register, and latency affects the GPU performance of WW3.

58 GEOSCIENCES↗

NASA NTRS2019

Multi-Node Modeling of Cryogenic Tank Pressurization System using Generalized Fluid System Simulation Program

This paper presents a multi-node model of autogenous pressurization of cryogenic propellant in a flight tank using the Generalized Fluid System Simulation Program (GFSSP), a general purpose flow network code developed at NASA/Marshall Space Flight Center. Tests were conducted to measure the pressure and temperatures at the various axial locations of the stratified ullage at 75% and 45% fill level. Liquid nitrogen was pressurized by gaseous nitrogen from a supply tank while the drain valve from the tank remained closed during the pressurization process. The ullage was discretized into 25 uniformly distributed nodes: 5 in the radial direction and 5 in the axial direction assuming the flow to be axisymmetric. Heat and mass transfer between the liquid and vapor has been modeled at the liquid vapor interface. Heat transfer between wall and vapor at the ullage has been accounted for by assuming heat transfer occurs by natural convection. The model also accounts for heat leak to the tank through the insulation and metal wall by heat conduction. The predicted pressures and temperatures are compared with the measured data.

Bandyopadhyay, Alak↗

DOE OSTI2021

Accelerating Multigrid-based Hierarchical Scientific Data Refactoring on GPUs

Rapid growth in scientific data and a widening gap between computational speed and I/O bandwidth make it increasingly infeasible to store and share all data produced by scientific simulations. Instead, we need methods for reducing data volumes: ideally, methods that can scale data volumes adaptively so as to enable negotiation of performance and fidelity tradeoffs in different situations. Multigrid-based hierarchical data representations hold promise as a solution to this problem, allowing for flexible conversion between different fidelities so that, for example, data can be created at high fidelity and then transferred or stored at lower fidelity via logically simple and mathematically sound operations. However, the effective use of such representations has been hindered until now by the relatively high costs of creating, accessing, reducing, and otherwise operating on such representations. We describe here highly optimized data refactoring kernels for GPU accelerators that enable efficient creation and manipulation of data in multigrid-based hierarchical forms. We demonstrate that our optimized design can achieve up to 250 TB/s aggregated data refactoring throughput—83% of theoretical peak—on 1024 nodes of the Summit supercomputer. We showcase our optimized design by applying it to a large-scale scientific visualization workflow and the MGARD lossy compression software.

Data refactoring↗

DOE OSTI2021

Accelerating Multigrid-based Hierarchical Scientific Data Refactoring on GPUs

Rapid growth in scientific data and a widening gap between computational speed and I/O bandwidth make it increasingly infeasible to store and share all data produced by scientific simulations. Instead, we need methods for reducing data volumes: ideally, methods that can scale data volumes adaptively so as to enable negotiation of performance and fidelity tradeoffs in different situations. Multigrid-based hierarchical data representations hold promise as a solution to this problem, allowing for flexible conversion between different fidelities so that, for example, data can be created at high fidelity and then transferred or stored at lower fidelity via logically simple and mathematically sound operations. However, the effective use of such representations has been hindered until now by the relatively high costs of creating, accessing, reducing, and otherwise operating on such representations. We describe here highly optimized data refactoring kernels for GPU accelerators that enable efficient creation and manipulation of data in multigrid-based hierarchical forms. We demonstrate that our optimized design can achieve up to 250 TB/s aggregated data refactoring throughput—83% of theoretical peak—on 1024 nodes of the Summit supercomputer. We showcase our optimized design by applying it to a large-scale scientific visualization workflow and the MGARD lossy compression software.

Chen, Jieyang↗

NASA NTRS2016

Multi-Disciplinary, Multi-Fidelity Discrete Data Transfer Using Degenerate Geometry Forms

In a typical multi-fidelity design process, different levels of geometric abstraction are used for different analysis methods, and transitioning from one phase of design to the next often requires a complete re-creation of the geometry. To maintain consistency between lower-order and higher-order analysis results, Vehicle Sketch Pad (OpenVSP) recently introduced the ability to generate and export several degenerate forms of the geometry, representing the type of abstraction required to perform low- to medium-order analysis for a range of aeronautical disciplines. In this research, the functionality of these degenerate models was extended, so that in addition to serving as repositories for the geometric information that is required as input to an analysis, the degenerate models can also store the results of that analysis mapped back onto the geometric nodes. At the same time, the results are also mapped indirectly onto the nodes of lower-order degenerate models using a process called aggregation, and onto higher-order models using a process called disaggregation. The mapped analysis results are available for use by any subsequent analysis in an integrated design and analysis process. A simple multi-fidelity analysis process for a single-aisle subsonic transport aircraft is used as an example case to demonstrate the value of the approach.

Olson, Erik D.↗

NASA NTRSDate not supplied

Two-Dimensional Network Flow Modeling of No-Vent Tank Filling of A Cryogenic Tank With Thermo-Dynamic Vent System Assisted Injector

A finite volume-based network flow modeling tool, Generalized Fluid System Simulation Program (GFSSP), was used to simulate the no-vent filling of a cryogenic tank with a thermodynamic vent system assisted injector. The tank was discretized into nodes in the radial and axial directions to calculate two-dimensional axisymmetric flow with heat transfer between the solid wall and fluid. The nodes are connected by branches where momentum equations are solved to calculate the flowrates. The mass and energy conservation equations and the equation of state are solved to calculate pressure, temperature, and resident mass at the nodes. The energy conservation equations at the solid nodes were solved to calculate the temperature of the solid wall. The system of equations was solved by a combination of successive substitution and simultaneous Newton-Raphson method. The numerical model accounts for a) different regimes of pool boiling heat transfer, b) condensation of vapor around the spray droplets, c) condensation of vapor at the interface of the cooled injector and ullage, and d) condensation at the liquid-vapor interphase. The numerical predictions of tank pressure, filling rate, and wall temperatures were compared with the test data. The paper also presents the effect of nodal discretization by comparing the results of the 1D and 2D models and examines the effect of the droplet diameter on tank filling.

Nodal Model↗

NASA NTRS2023

Two-Dimensional Network Flow Modeling of No-Vent Tank Filling of A Cryogenic Tank With Thermo-Dynamic Vent System Assisted Injector

A finite volume-based network flow modeling tool, Generalized Fluid System Simulation Program (GFSSP), was used to simulate the no-vent filling of a cryogenic tank with a thermodynamic vent system assisted injector. The tank was discretized into nodes in the radial and axial directions to calculate two-dimensional axisymmetric flow with heat transfer between the solid wall and fluid. The nodes are connected by branches where momentum equations are solved to calculate the flowrates. The mass and energy conservation equations and the equation of state are solved to calculate pressure, temperature, and resident mass at the nodes. The energy conservation equations at the solid nodes were solved to calculate the temperature of the solid wall. The system of equations was solved by a combination of successive substitution and simultaneous Newton-Raphson method. The numerical model accounts for a) different regimes of pool boiling heat transfer, b) condensation of vapor around the spray droplets, c) condensation of vapor at the interface of the cooled injector and ullage, and d) condensation at the liquid-vapor interphase. The numerical predictions of tank pressure, filling rate, and wall temperatures were compared with the test data. The paper also presents the effect of nodal discretization by comparing the results of the 1D and 2D models and examines the effect of the droplet diameter on tank filling.

Nodal Model↗

DOE OSTI2023

Porting fragmentation methods to GPUs using an OpenMP API: Offloading the resolution-of-the-identity second-order Møller–Plesset perturbation method

Here, using an OpenMP Application Programming Interface, the resolution-of-the-identity second-order Møller–Plesset perturbation (RI-MP2) method has been off-loaded onto graphical processing units (GPUs), both as a standalone method in the GAMESS electronic structure program and as an electron correlation energy component in the effective fragment molecular orbital (EFMO) framework. First, a new scheme has been proposed to maximize data digestion on GPUs that subsequently linearizes data transfer from central processing units (CPUs) to GPUs. Second, the GAMESS Fortran code has been interfaced with GPU numerical libraries (e.g., NVIDIA cuBLAS and cuSOLVER) for efficient matrix operations (e.g., matrix multiplication, matrix decomposition, and matrix inversion). The standalone GPU RI-MP2 code shows an increasing speedup of up to 7.5× using one NVIDIA V100 GPU with one IBM 42-core P9 CPU for calculations on fullerenes of increasing size from 40 to 260 carbon atoms using the 6-31G(d)/cc-pVDZ-RI basis sets. A single Summit node with six V100s can compute the RI-MP2 correlation energy of a cluster of 175 water molecules using the correlation consistent basis sets cc-pVDZ/cc-pVDZ-RI containing 4375 atomic orbitals and 14 700 auxiliary basis functions in ~0.85 h. In the EFMO framework, the GPU RI-MP2 component shows near linear scaling for a large number of V100s when computing the energy of an 1800-atom mesoporous silica nanoparticle in a bath of 4000 water molecules. The parallel efficiencies of the GPU RI-MP2 component with 2304 and 4608 V100s are 98.0% and 96.1%, respectively.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

NASA NTRS2008

Impacts of the IBM Cell Processor to Support Climate Models

NASA is interested in the performance and cost benefits for adapting its applications to the IBM Cell processor. However, its 256KB local memory per SPE and the new communication mechanism, make it very challenging to port an application. We selected the solar radiation component of the NASA GEOS-5 climate model, which: (1) is representative of column physics (approximately 50% computational time), (2) has a high computational load relative to transferring data from and to main memory, (3) performs independent calculations across multiple columns. We converted the baseline code (single-precision, Fortran) to C and ported it with manually SIMDizing 4 independent columns and found that a Cell with 8 SPEs can process 2274 columns per second. Compared with the baseline results, the Cell is approximately 5.2X, approximately 8.2X, approximately 15.1X faster than a core on Intel Woodcrest, Dempsey, and Itanium2, respectively. We believe this dramatic performance improvement makes a hybrid cluster with Cell and traditional nodes competitive.

Zhou, Shujia↗

NASA NTRS1985

Introduction to the Space Physics Analysis Network (SPAN)

The Space Physics Analysis Network or SPAN is emerging as a viable method for solving an immediate communication problem for the space scientist. SPAN provides low-rate communication capability with co-investigators and colleagues, and access to space science data bases and computational facilities. The SPAN utilizes up-to-date hardware and software for computer-to-computer communications allowing binary file transfer and remote log-on capability to over 25 nationwide space science computer systems. SPAN is not discipline or mission dependent with participation from scientists in such fields as magnetospheric, ionospheric, planetary, and solar physics. Basic information on the network and its use are provided. It is anticipated that SPAN will grow rapidly over the next few years, not only from the standpoint of more network nodes, but as scientists become more proficient in the use of telescience, more capability will be needed to satisfy the demands.

Green, J. L.↗

NASA NTRS2020

Model Validation for Bigelow Expandable Activity Module (BEAM) with Stowage

The Bigelow Expandable Activity Module (BEAM) was a collaboration between NASA and Bigelow Aerospace to develop an expandable habitat technology that could be used for future space exploration missions such as the Artemis program. On April 26, 2016, the BEAM was launched and berthed to the International Space Station (ISS) on the aft port of Node 3. After the originally planned two-year mission life, BEAM was approved for use as a stowage module for a 109-cargotransfer bag equivalent (CTBE) to alleviate stowage limitations onboard the ISS. The present work builds on the previous 2017 model validation to flight data that demonstrated that BEAM could generally accommodate cargo stowage without adversely affecting BEAM thermal control, primarily for condensation management. Once the stowage was configured in early 2019, model validation efforts resumed. The thermal model revisions included the CTBE stowage and updated CFD-derived heat transfer coefficients for various intra-module ventilation (IMV) flow rates. The coefficients were provided by the Crew and Thermal Systems Division (EC) at NASA JSC and Jacobs Technology. The model is now revalidated and able to predict minimum temperatures with good agreement to flight temperatures (within 1 °C) for both historic and recurring minimum values. Additionally, the model is partially validated for reduced flow rates. With this model, the team was able to provide new flight rule recommendations for condensation and IMV operations management.

Sydney Taylor↗