Search NASA⌕ Search

SEARCH · Search NASA

Results for “program architectures”

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 955 records · Page 53

National Energy Education Development Project (NEED Project) (CRADA Final Report)

The U.S. Department of Energy Building Technologies Office (BTO) funds student competitions that introduce students to careers in the building sciences and increase public awareness around high-performance buildings to support the goal of developing, demonstrating, and accelerating the adoption of cost-effective technologies, techniques, tools, and services that enable high-performing, energy-efficient and demand-flexible residential and commercial buildings in both the new and existing buildings markets. The U.S. Department of Energy Solar Decathlon® (DOE/SD) is a flagship, high-visibility international competition started in 2002 that advances the goals of BTO by introducing students to building science careers; educating students and the public about the latest technologies and materials in high-performance buildings; encouraging student-led projects and research centered around building science; and demonstrating to the public the comfort and savings of homes that combine energy-efficient construction, home systems, appliances and innovative design with onsite renewable energy production. SD is a collegiate competition, comprising 10 contests, that challenges student teams to design and build highly efficient and innovative buildings powered by renewable energy. The winners will be those teams that best blend architectural and engineering excellence with innovation, market potential, building efficiency, and smart energy production. Solar Decathlon is comprised of two Challenges – Design Challenge (annual) and Build Challenge (biennial). The National Renewable Energy Laboratory (NREL) provides competition management for Solar Decathlon. NREL and Participant establish this CRADA to enable the success of the overall Solar Decathlon program by managing sponsorship funds and creating a K12 education program. Participant is to act as an Education Partner to Solar Decathlon, which includes: 1) accepting and dispersing sponsorship funds for DOE/SD; and 2) providing K12 education program to support Solar Decathlon Competition Events in April each year.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

GraphReduce: Processing Large-Scale Graphs on Accelerator-Based Systems

Recent work on real-world graph analytics has sought to leverage the massive amount of parallelism offered by GPU devices, but challenges remain due to the inherent irregularity of graph algorithms and limitations in GPU-resident memory for storing large graphs. We present GraphReduce, a highly efficient and scalable GPU-based framework that operates on graphs that exceed the device’s internal memory capacity. GraphReduce adopts a combination of edge- and vertex-centric implementations of the Gather-Apply-Scatter programming model and operates on multiple asynchronous GPU streams to fully exploit the high degrees of parallelism in GPUs with efficient graph data movement between the host and device.

Architecture optimization, reuse, performance, ene↗

Review of TEAMER Awards for WEC-Sim Support: Preprint

Testing Expertise and Access for Marine Energy Research (TEAMER) is a U.S. Department of Energy Water Power Technologies Office sponsored program, overseen by the Pacific Ocean Energy Trust, which aims to advance the state of marine energy technologies. The program connects technology developers with experts at U.S. facilities, including numerical modeling and analysis facilities, like WEC-Sim. The WEC-Sim facility is supported by the WEC-Sim development team at Sandia National Laboratories and the National Renewable Energy Laboratory. WEC-Sim (Wave Energy Converter SIMulator) is an open-source software for simulating wave energy converters. WEC-Sim can model the multi-body dynamics of devices comprised of bodies, joints, power take-off systems, and mooring systems. Since TEAMER's first round of support in 2020, there have been eighteen TEAMER awards focused on numerical model development in WEC-Sim. TEAMER awards for WEC-Sim support have modeled a wide range of wave energy converter archetypes, including point absorbers, attenuators, oscillating water columns, and many other novel architectures. A wide variety of studies have been conducted, leading to important insights for TEAMER partners and software improvements for WECSim. This article highlights several successful WEC-Sim TEAMER awards. The awards described herein include TEAMER recipients Ocean Motion Technologies, AquaHarmonics, iProTech, East Carolina University, Virginia Tech, Maiden Wave Energy, and the University of Massachusetts Dartmouth. The awards of these seven partners contain a wide range of investigations and cover the creation of baseline hydrodynamic models, PTO modeling, geometry optimization in both boundary element methods and WECSim, and model tuning and validation.

industry support↗

Tail queues: A multi-threaded matching architecture

As we approach exascale, computational parallelism will have to drastically increase in order to meet throughput targets. Many–core architectures have exacerbated this problem by trading reduced clock speeds, core complexity, and computation throughput for increasing parallelism. This presents two major challenges for communication libraries such as MPI: the library must leverage the performance advantages of thread level parallelism and avoid the scalability problems associated with increasing the number of processes to that scale. Hybrid programming models, such as MPI+X, have been proposed to address these challenges. MPI THREAD MULTIPLE is MPI's thread safe mode. While there has been work to optimize it, it largely remains non–performant in most implementations. While current applications avoid MPI multithreading due to performance concerns, it is expected to be utilized in future applications. One of the major synchronous data structures required by MPI is the matching engine. In this paper, we present a parallel matching algorithm that can improve MPI matching for multithreaded applications. We then perform a feasibility study to demonstrate the performance benefit of the technique.

97 MATHEMATICS AND COMPUTING↗

The SONATA data format for efficient description of large-scale network models

Increasing availability of comprehensive experimental datasets and of high-performance computing resources are driving rapid growth in scale, complexity, and biological realism of computational models in neuroscience. To support construction and simulation, as well as sharing of such large-scale models, a broadly applicable, flexible, and high-performance data format is necessary. To address this need, we have developed the Scalable Open Network Architecture TemplAte (SONATA) data format. It is designed for memory and computational efficiency and works across multiple platforms. The format represents neuronal circuits and simulation inputs and outputs via standardized files and provides much flexibility for adding new conventions or extensions. SONATA is used in multiple modeling and visualization tools, and we also provide reference Application Programming Interfaces and model examples to catalyze further adoption. SONATA format is free and open for the community to use and build upon with the goal of enabling efficient model building, sharing, and reproducibility.

59 BASIC BIOLOGICAL SCIENCES↗

Cybersecurity for the Operational Technology Environment (CyOTE) (Final Technical Report)

Electric grids have historically been susceptible to both physical attacks and environmental hazards but the implementation of smart grids, remote management, and self-healing networks, has now made the grid vulnerable to cyber attacks. To address risks introduced by routable connectivity, utilities must establish dynamic solutions to identify, protect, detect, respond to, and recover from cyber security threats and vulnerabilities. In response to the evolving threat landscape U.S. Department of Energy-Office of Cybersecurity, Energy Security, and Emergency Response (DOE CESER) initiated the Cybersecurity for the OT Environment (CyOTE) pilot program, a U.S. Department of Energy (DOE) effort designed to leverage U.S. intelligence capabilities to prevent, detect, or mitigate a cyber attack on utility operational technology (OT) networks. As part of the CyOTE pilot, The Southern Company (Southern Company or Southern) researched, evaluated and deployed emerging Commercial off the Shelf (COTS) technologies and cyber security monitoring architectures to provide previously unrealized network visibility and situational awareness through deep packet inspection and data analytics. This Final Scientific/Technical Report documents the objectives, methodology, lessons learned, and results of Southern Company’s participation in the CyOTE pilot from December 2018 to September 2023.

24 POWER TRANSMISSION AND DISTRIBUTION↗

High Temperature Additive Architectures for 65% Efficiency (Final Technical Report)

This project aimed to develop advanced high-temperature additive components that contribute towards the DOE’s goal for advanced gas turbines that are capable of at least 65% efficiency in combined cycle application. The objective was to leverage state-of-the-art additive manufacturing to develop an innovative stage 1 turbine nozzle (S1N) that can provide cooling flow savings while maintaining the component durability expected in today’s gas turbines. The program had two phases. Phase I was a conceptual phase for novel advanced cooling designs enabled by additive manufacturing, as well as proposals for validation. Phase II included execution of the Phase I conceptual design, including manufacturing of prototype hardware and validation within an environment that is similar to engine operation. During Phase I of the program, the team devised a concept to reduce cooling air usage. The cooling air used in the side walls is filmed out along the side walls, while the cooling air used in the airfoil is eventually directed to near-wall channels and exits holes along the airfoil trailing end. During this program, the team performed additive trials to analyze the geometric limitations of additive manufacturing. This helped the team understand minimum wall thicknesses, hole sizes, and cooling channel dimensions among other limits. Phase II of this program pushed GE Vernova beyond its previous experience of designing and manufacturing an additively manufactured hot gas path component. Modern hot gas path components utilize material chemistries that are traditionally hard to weld, such as cast Renè 108, and exhibit solidification cracking when additively manufactured using Direct Metal Laser Melting (DMLM). Note that AM108 is a powder form of Renè 108. A S1N with advanced cooling is larger and more complex than parts previously built by additive manufacturing and required new learnings to resolve risks around solidification cracking. Finally, the team validated the design in a combustion rig that replicated operation in a gas turbine. In order to properly quantify the benefits of the new additive design, a baseline was also tested in the rig and operated under the same conditions. In addition, an uncertainty analysis was done to quantify any sources of error that could impact the results. At the end of the validation effort, it was determined that the additive nozzle exceeded the 15% reduction in cooling flow goal even with the worst-case assumptions for uncertainty.

03 NATURAL GAS↗

Decision Point 3 of Statement of Project Objectives (SOPO) “Recovery Act: Development of ITM Oxygen Technology for Integration with Advanced Industrial Systems”

Air Products is carrying out a scope of work under Phase 5 of the ITM Oxygen Cooperative Agreement to design, build, and operate a ceramic membrane fabrication facility (the -CerFabII) to enable production of membrane modules to supply a conceptual 2000 ton per day (TPD) ITM Oxygen facility (the -ITM Oxygen Development FacilityII), and to perform supporting development tasks in materials development and engineering development toward industrial, carbon capture and sequestration applications. Air Products is executing this project under the American Recovery and Reinvestment Act (ARRA) with the objective to accelerate the adoption of ITM Oxygen technology to help meet the country’s goals for deploying clean power plants. The objective of this Topical Report is to address the requirements of Decision Point 3 (DP3), which pertains to the status of all Tasks within Phase 5 and most notably the project status of the CerFab (Task 30) prior to authorization of funds for equipment purchase and construction of the facility. The intent of the DP3 is to provide the opportunity for DOE-NETL to review the status of these tasks and to make recommendations on forward project direction, including a recommendation to pass into Budget Period 8. In the area of Materials Development, Air Products has specified a high pressure dilatometer system which will enable measurements of material expansion of ITM ceramic compounds at very high oxygen partial pressures consistent with CCS applications. Under Task 28.2, subcontractor Ceramatec has made significant progress since DP2 in materials selection and process development and improvement for advanced architecture module fabrication. Ceramatec has determined a materials specification, and has selected a process for making the material. Ceramatec has further developed and selected the process for applying the membrane to unsintered advanced architecture wafers with a Two Step process. Ceramatec has built submodules meeting leak rate specifications and intact 1-TPD modules compromised of advanced architecture components. Equipment and processes required for implementation of the advanced architecture design in the CerFab have been identified and are included in the Task 30 scope of equipment supply. Under Task 29.0, Air Products has developed conceptual ideas for implementation of ITM Oxygen in a 2000 TPD test unit. These concepts are proposed for scale up of the ITM technology to large scale with reduced commercial risk. Under Task 30, Air Products and Ceramatec have collaborated to re-design the CerFab process with the advanced architecture wafer production process as primary, and laid out the new process in the Tooele building; the required process equipment have been specified. The project team updated the project cost and schedule estimates, including all environmental permit schedules. The team also re-assessed project risk and costs for the commissioning and operating phases of the program. Newly projected costs for the CerFab, for which details are provided here as attachments, indicate a savings of approximately $100,000 relative to the original estimate at the time of Phase 5 contract definitization. However, an assessment of risks associated with the equipment, process, and operations of the CerFab indicate that approximately $2.9 million of costs should be expected to address various issues in the project. In addition, $1.6 million of additional spending is expected relative to the original estimate to operate the facility during Task 30.2. Air Products proposes a plan to re-direct funds from Tasks 28, 29, and 31 to Task 30 as necessary to allow the CerFab to meet the Phase 5 program objectives.

Armstrong, Phillip↗

Developing And Scaling an OpenFOAM Model to Study Turbulent Flow in a HFIR Coolant Channel

Improving the understanding of how computational fluid dynamics (CFD) direct numerical simulations (DNS) of flows in the High Flux Isotope Reactor (HFIR) perform when run in parallel using the high performance computing (HPC) platform Summit at the Oak Ridge Leadership Computing Facility (OLCF) is of particular importance to boost the computational tools used to support HFIR conversion to low enriched fuel (LEU). Evaluation of scaling performance was driven by the increasing importance of graphics processing unit (GPU) usage in HPC, which is becoming the standard for modern supercomputers such as Summit. The desired results are to obtain a strong positive correlation between the computational resources dedicated to a problem and the relative speed-up of the simulation in comparison to a benchmark. This capability will allow substantially improvement in HFIR flow analytical capabilities, specifically when predicting turbulence properties at high Reynolds numbers. The study leverages previous simulation results performed with code PHASTA (finite element) on HPC platforms Cori (NERSC) and Theta (ALCF) [1] with computing options provided in the computing platform OpenFOAM (finite volume) at OLCF. Transitioning from PHASTA to OpenFOAM will (1) eliminate dependence on third-party software for mesh generation and manipulation, (2) reduce resource needs by employing modern architectures, and (3) build expertise for future modeling of HFIR-specific problems like heat transfer in involute geometry, entrance effects, flow structure in channel corners, and so on—all important issues when defining the available thermal margins in the transition to LEU. CPUs and GPUs differ significantly in their architecture and utilization, as discussed in the literature [2]. The most important differences are in the approach to computations and their memory. A single GPU contains a large quantity of cores, enabling it to perform with a much higher throughput than a CPU, but execution requires a different approach. GPU codes execute instructions using the Single-Instruction Multiple-Thread (SIMT) approach in which a single instruction is used for groups of threads called warps. A warp typically consists of 32 threads which must execute the same set of instructions, although on separate threads. Alternately, a CPU has far fewer cores that are much more flexible in their operation, excelling at quickly performing more complex serial computations. This is why GPUs have greater throughput when properly utilized. The second important difference is seen when comparing their memory spaces. Limited memory allocations and CPU–GPU communications cause a significant bottleneck in GPU-accelerated programs. Further study was required to properly take advantage of GPU resources. A comprehensive analysis of code performance and the model-specific features of turbulence constitutes the core of this work. In this study, a DNS simulation of HFIR channel turbulence was performed with the finite volume CFD code OpenFOAM v2112 and CUDA v11.0 on Red Hat Enterprise Linux v8.2. The OpenFOAM installation had AMGx integrated to enable GPU acceleration and utilizes the PETSc4FOAM library. The computational resources and the problem size were scaled on CPU and CPU + GPU architectures to gain a better understanding of the performance of a DNS problem on modern computing hardware. The study aimed to analyze the scaling of the code exclusively on CPUs and then to examine the scaling of the codes with GPU acceleration enabled. Scaling studies included CPU and GPU acceleration on a mesh of varying resolution to analyze the impact of problem size relative to computational resources. In the course of preparing the GPU configuration on Summit, mainly using the AMGX solvers, difficulties were encountered stemming from constant changes resulting from extensive ongoing development activities and the changing environment. This resulted in the inability to complete the GPU portion of the work. The code was compiled and tested, but production runs to assess acceleration were not performed because the used discretional compute time allocation expired as year-end approached. The Summit HPC platform is scheduled for decommissioning in 2024, making it unattractive for future use with Nvidia-based GPUs. Therefore, the work will be moved onto NERSC machines in FY24. An application was prepared and submitted, and sufficient node-hours were awarded to continue the research in the next calendar year. This report summarizes work performed thus far, which mostly focused on CPU OpenFOAM computing.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Hierarchical Composites Patterned via 3D Printed Cellular Fluidics

Additive manufacturing of freeform structures containing multiple materials with deterministic spatial arrangement and interactions remains a challenge for most 3D printing processes, due to complex fabrication tool requirements and limitations in printability of some material classes. Here, in this paper, a versatile method is reported to produce architected composites using the concept of cellular fluidics, in which lattices of unit cells are used as templating scaffolds to guide flowable infill materials in a programmed spatial pattern, upon which they are cured in place to produce a deterministically ordered multimaterial solid. The lattice design relies on the unit cell size, type, strut diameter, surface wetting, and distribution of cellular structures to control liquid flow and retention. Individual unit cells are tuned to achieve reliable infilling and combined into higher-order architectures to achieve multiscale composite materials with disparate mechanical properties, including those considered non-printable. Lattice design considerations for leveraging capillary phenomena and demonstrate several methods of patterning polymers in 3D-printed cellular fluidic structures are presented. The concept of tuning the compressive response of an architected composite using a flexible-elastomer as the lattice and a stiff-epoxy as the infill material is illustrated.

36 MATERIALS SCIENCE↗

Human readiness levels and Human Views as tools for user-centered design

The Human Readiness Level (HRL) scale is a simple nine-level scale that brings structure and consistency to the real-world application of user-centered design. It enables multidisciplinary consideration of human-focused elements during the system development process. Use of the standardized set of questions comprising the HRL scale results in a single human readiness number that communicates system readiness for human use. The Human Views (HVs) are part of an architecture framework that provides a repository for human-focused system information that can be used during system development to support the evaluation of HRL levels. Here, this paper illustrates how HRLs and HVs can be used in combination to support user-centered design processes. A real-world example for a U.S. Army software modernization program is described to demonstrate application of HRLs and HVs in the context of user-centered design.

42 ENGINEERING↗

Supporting the Global Arrays PGAS Model Using MPI One-Sided Communication

The industry-standard Message Passing Interface (MPI) provides one-sided communication functionality and is available on virtually every parallel computing system. However, it is believed that MPI’s one-sided model is not rich enough to support higher-level global address space parallel programming models. We present the first successful application of MPI one-sided communication as a runtime system for a PGAS model, Global Arrays (GA). This work has an immediate impact on users of GA applications, such as NWChem, who often must wait several months to a year or more before GA becomes available on a new architecture. We explore challenges present in the application of MPI-2 to PGAS models and motivate new features in the upcoming MPI-3 standard. The performance of our system is evaluated on several popular high-performance computing architectures through communication benchmarking and application benchmarking using the NWChem computational chemistry suite.

ARMCI↗

GraphReduce: Large-Scale Graph Analytics on Accelerator-Based HPC Systems

Recent work on real-world graph analytics has sought to leverage the massive amount of parallelism offered by GPU devices, but challenges remain due to the inherent irregularity of graph algorithms and limitations in GPU-resident memory for storing large graphs. We present GraphReduce, a highly efficient and scalable GPU-based framework that operates on graphs that exceed the device’s internal memory capacity. GraphReduce adopts a combination of both edge- and vertex-centric implementations of the Gather-Apply-Scatter programming model and operates on multiple asynchronous GPU streams to fully exploit the high degrees of parallelism in GPUs with efficient graph data movement between the host and the device.

Architecture optimization, reuse, performance, ene↗

Large Scale Computing and Storage Requirements for Basic Energy Sciences: Target 2017. Report of the NERSC Requirements Review Conducted October 8–9, 2013, Gathersburg, MD

The National Energy Research Scientific Computing Center (NERSC) is the primary computing center for the DOE Office of Science, serving approximately 5,000 users working on some 700 projects that involve nearly 600 codes in a wide variety of scientific disciplines. In addition to large-scale computing and storage resources NERSC provides support and expertise that help scientists make efficient use of its systems. The latest review revealed several key requirements, in addition to achieving its goal of characterizing BES computing and storage needs. high-level findings are: 1) Scientists will need access to significantly more computational and storage resources to achieve their goalds and reach BES research objectives; 2) Users will need assistance from NERSC to prepare for Cori (NERSC-8) and follow-on manycore systems; 3) Research teams need to run complex jobs of many different types and scales; 4) BES is a leader in innovative use of HPC and requires a diverse set of resources and services from NERSC; and 5) BES facilities need computational analysis and data storage resources beyond what they can provide. This report expands upon these key points and add others. The results are based upon representative samples, called "case studies," of the needs of science teams within BES. The case study topics were selected by the NERSC meeting coordinators and BES program managers to represent the BES production computing workload. Prepared by BES workshop participants, the case studies contain a summary of science goals, methods of solution, current and future computing requirements, and special software and support needs. Also included are strategies for computing in the highly parallel "many-core" environment that is expected to dominate HPC architectures over the next few years.

97 MATHEMATICS AND COMPUTING↗

Data Locality Enhancement of Dynamic Simulations for Exascale Computing (Final Report)

The development of modern processors exhibits two trends that complicate the optimizations of modern software. The first is the increasing sensitivity of processors' throughput to irregularities in computation. With more processors produced through a massive integration of simple cores, future systems will increasingly favor regular data-level parallel computations, but deviate from the needs of applications with complex patterns. Some evidences are already shown on Graphic Processing Units (GPU): Irregular data accesses (e.g., indirect references A[D[i]]) and conditional branches are limiting many GPU applications' performance at a level an order of magnitude lower than the peak of GPU. The second hardware trend is the growing gap between memory bandwidth and the aggregate speed—that is, the sum of all cores' computing power—of a Chip Multiprocessor (CMP). Despite the capped growth of the peak CPU speed, the aggregate speed of a CMP keeps increasing as more cores get into a single chip. It is expected that by 2018, node concurrency in an exascale system will increase by hundreds of times, whereas, memory bandwidth will expand by only 10 to 20 times. Consequently, data movement and storage is expected to consume more than 70% of the total system power. Bridging this gap is difficult; the complexities of modern CMP memory hierarchy make it even harder: Data cache becomes shared among computing units, and the sharing is often non-uniform—whether two computing units share a cache depends on their proximity and the level of the cache. On the recent IBM Power7 architecture, for instance, four hardware contexts (or SMT threads) in a core share the entire memory hierarchy, all cores in one chip share an on-chip L3 cache, and cores across chips share L3 and main memory through off-chip connections. These two trends complicate the translation of computing power into performance, especially for a program with either intensive data accesses or complex patterns in data accesses or control flow paths. Unfortunately, both attributes present and will persist in a class of important applications. For instance, many scientific simulations deal with a large volume of data. And meanwhile, as most real-world processes are non-uniform and evolving (e.g., the evolution of a galaxy or the process of a drug injection), both the computations and data accesses of these programs tend to be irregular and dynamically changing. Currently, the lack of support to these applications on modern CMP severely limits their performance. On GPU, as our recent study shows and other studies echo, performance enhancement of a factor of integers is possible when memory accesses or control flows are streamlined for a set of GPU applications. On multicore CPU, our studies show that traditional locality enhancement, for being oblivious to the new features of multicore memory hierarchy, may even cause large slowdown to data-intensive dynamic applications. The severity of the issues is expected to worsen as the two hardware trends continue. Some recent studies try to match software with the trends, but in a limited scope or manner. For irregularities on GPU, most studies focus on irregularities analyzable through static analysis (e.g., data accesses in regular loops). Dynamic irregularities are harder to address because the needed analysis and transformations typically have to happen at run time. Some other research resorts to hardware extensions, an actual adoption of which is unclear for the entailed space cost and complexity. For data locality, recent years have seen some exploitations of the new memory hierarchy on multicore for performance, but most of them are on process or thread scheduling, rather than program transformations. Our recent study reveals that program-level transformations may magnify the scheduling benefits by a factor of seven, concluding that program-level transformation should play a central role for data locality enhancement on modern CMP. But research in this direction has been sparse, and most have focused on data layout or cache performance modeling, rather than program transformations to match with the new memory hierarchy features. Overall, it is still an open question how to bridge the gap between dynamic computations and the two prominent properties of modern processors. The goal of this project is to develop a set of techniques and software tools to enhance the matching between memory accesses in dynamic simulations and the prominent features of modern and future CMP, alleviating the memory performance issues for petascale and exascale computing. This report summarizes the discoveries and products produced throughout this project. It includes free launch, a new software approach to overcoming the shortcomings of both methods; coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution; algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Supporting dynamic parallelism is important for GPU to benefit a broad range of applications. There are currently two fundamental ways for programs to exploit dynamic parallelism on GPU: a software-based approach with software-managed worklists, and a hardware-based approach through dynamic subkernel launches. Neither is satisfactory. The former is complicated to program and is often subject to some load imbalance; the latter suffers large runtime overhead. In this work, we propose free launch, a new software approach to overcoming the shortcomings of both methods. It allows programmers to use subkernel launches to express dynamic parallelism. It employs a novel compiler-based code transformation named subkernel launch removal to replace the subkernel launches with the reuse of parent threads. Coupled with an adaptive task assignment mechanism, the transformation reassigns the tasks in the subkernels to the parent threads with a good load balance. The technique requires no hardware extensions, immediately deployable on existing GPUs. It keeps the programming convenience of the subkernel launch-based approach while avoiding its large runtime overhead. Meanwhile, its superior load balancing makes it outperform manual worklist-based techniques by 3X on average. The work was published at the 48th Annual IEEE/ACM International Symposium on Microarchitecture (Micro'2015). A Graphic Processing Unit (GPU) system is typically equipped with many types of memory (e.g., global, constant, texture, shared, cache). Data placement determines what data are placed on which type of memory, essential for GPU memory performance. Prior optimizations of data placement always require a single view of a data object on memory, which limits the optimization effectiveness. In this work, we propose coherence-free multiview, an approach that allows multiple views of a single data object to co-exist on GPU memory during a GPU kernel execution. We demonstrate that under certain conditions, the multiple views can remain incoherent while facilitating enhanced data placement. We present a theorem and some compiler support to ensure the soundness of the usage of coherence-free multiview. We further develop reference-discerning data placement, a new way to enhance data placements on GPU. It enables more flexible data placements by using coherence-free multiview to leverage the slack in coherence requirement of some GPU programs. Experiments on three types of GPU systems show that, with less than 200KB space cost, the new data placement technique can provide a 1.6X average (up to 4.27X) speedup. In addition, we have examined algorithmic optimizations to data analytics problems, especially those that involve lots of distance calculations. Computing distances among data points is an essential part of many important algorithms in data analytics, graph analysis, and other domains. In each of these domains, developers have spent significant manual effort optimizing algorithms, often through novel applications of the triangle equality, in order to minimize the number of distance computations in the algorithms. In this work, we observe that many algorithms across these domains can be generalized as an instance of a generic distance-related abstraction. Based on this abstraction, we derive seven principles for correctly applying the triangular inequality to optimize distance-related algorithms. Guided by the findings, we develop Triangular OPtimizer (TOP), the first software framework that is able to automatically produce optimized algorithms that either matches or outperforms manually designed algorithms for solving distance-related problems. TOP achieves up to 237x speedups and 2.5X on average. The work has been published at the 32nd International Conference on Machine Learning (ICML'15) and the 41st International Conference on Very Large Data Bases (VLDB'15).

42 ENGINEERING↗

Authentication Protocol for ICS without Encryption

Industrial control systems (ICS) describes the different types of control systems and associated instrumentation, which includes the devices, systems, networks, and controls used to operate and/or automate industrial processes. Depending on the industry, each ICS functions differently and are built to electronically manage tasks efficiently. Historically, industrial plants have included networking, and changes have been made over time without consideration for security. Many ICS devices currently have inadequate security features. Adding encryption in ICS is undesirable because many of the new detection method involves deep packet inspection. Therefore, we propose to develop a method of providing authentication between devices without encrypting the data. This testbed design will allow rapid development and experimentation with Cyber Physical Systems. Modbus connects to a supervisory computer with a Remote Terminal Unit (RTU) in Supervisory Control and Data Acquisition (SCADA) systems. Our project seeks to enhance security in the ICS environment by adding device authentication to the lower levels of the plant floor. This approach was selected as it allows verification of the sender of data from level 1/level 0 devices up to supervisory systems. This addition is desirable as it adds enhanced verification to the integrity of system data without obscuring (via encryption) the data being transmitted across the network. This is important because many ICS systems are real time critical (decryption adds to latency) and many other good security options require deep packet inspection (which encryption completely defeats). To accomplish this, a proof of concept of adding an authentication scheme was designed and will be tested with a common ICS protocol. The setup includes a single serial cable connecting the serial ports on two devices, a Master and a Minion. Authentication will occur in the following methods: - Data layer involving Transport Layer Security/Secure Socket Layer (TLS/SSL); - TCP/IP route implementation; - Key exchange algorithm; - Raspberry Pi programming for authentication. A portion of the setup includes adding Raspberry Pi into the system as the device is handling the authentication for the network traffic. Modbus is the chosen ICS protocol. It is used to transmit signals from instrumentation and control devices back to a main controller or data gathering system. Typically, the device identity is assumed but not verified. For verifying device identity, we constructed a digital signature architecture with TLS. A novel adaptation of this is to use only the digital signature that is provided and not the encryption capability. This allows ICS traffic to flow unobscured while still providing verification for the integrity of the data. Verification of this process will involve adding additional I/O level network devices with spoofed IP addresses and detecting the authenticated versus unauthenticated traffic. Our TLS implementation will have encryption disabled but should still allow for digital signing of data. This will enable our system to distinguish between real end devices and spoofed end devices. This experiment is presently building the test setup. It is focusing on a small setting to identify authenticated and unauthenticated devices on the network. This should allow for scaling to a larger network size after the initial proof of concept. One concern with this scheme is a certificate deployment has a large scale key management. Network size will thus increase the key management task for security. In addition, we believe that further research should be done on exploring the use of this method to defeat man-in-the-middle attacks.

42 ENGINEERING↗

Architecture and operation of the Z Pulsed Power Facility vacuum system.

The Z Pulsed Power Facility at Sandia National Laboratories in Albuquerque, New Mexico, USA is one of the world's premier high energy density physics facilities. The Z Facility derives its name from the z-pinch phenomena which is a type of plasma confinement system that uses the electrical current in the plasma to generate a magnetic field that compresses it. Z refers to the direction of current flow, the z axis in a three dimensional Cartesian coordinate system. The multiterawatt, multimegajoule electrical pulse the Facility produces is 100-400 nanoseconds in time. Research and development programs currently being conducted on the Z Facility include inertial confinement fusion, dynamic material properties, laboratory astrophysics and radiation effects. The Z Facility vacuum system consists of two subsystems, center section and load diagnostics. Dry roughing pumps and cryogenic high vacuum pumps are used to evacuate the 40,000 liter, 200 square meter center section of the facility where the experimental load is located. Pumping times on the order of two hours are required to reduce the pressure from atmospheric to 10{sup -5} Torr. The center section is cycled from atmosphere to high vacuum for each experiment. The facility is capable of conducting one to two experiments per day. Numerous smaller vacuum pumping systems are used to evacuate load diagnostics. The megajoules of energy released during an experiment causes damage to the Facility that presents numerous challenges for reliable operation of the vacuum system.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

On the Design, Development, and Analysis of Optimized Matrix-Vector Multiplication Routines for Coprocessors

The manycore paradigm shift, and the resulting change in modern computer architectures, has made the development of optimal numerical routines extremely challenging. In this work, we target the development of numerical algorithms and implementations for Xeon Phi coprocessor architecture designs. In particular, we examine and optimize the general and symmetric matrix-vector multiplication routines (gemv/symv), which are some of the most heavily used linear algebra kernels in many important engineering and physics applications. We describe a successful approach on how to address the challenges for this problem, starting with our algorithm design, performance analysis and programing model and moving to kernel optimization. Our goal, by targeting low-level and easy to understand fundamental kernels, is to develop new optimization strategies that can be effective elsewhere for use on manycore coprocessors, and to show significant performance improvements compared to existing state-of-the-art implementations. Therefore, in addition to the new optimization strategies, analysis, and optimal performance results, we finally present the significance of using these routines/strategies to accelerate higher-level numerical algorithms for the eigenvalue problem (EVP) and the singular value decomposition (SVD) that by themselves are foundational for many important applications.

Kabir, Khairul↗