Search NASA⌕ Search

SEARCH · Search NASA

Results for “Computer software and hardware”

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 883 records · Page 49

The application of neural networks with artificial intelligence technique in the modeling of industrial processes

Neural networks are a relatively new artificial intelligence technique that emulates the behavior of biological neural systems in digital software or hardware. These networks can 'learn', automatically, complex relationships among data. This feature makes the technique very useful in modeling processes for which mathematical modeling is difficult or impossible. The work described here outlines some examples of the application of neural networks with artificial intelligence technique in the modeling of industrial processes.

99 GENERAL AND MISCELLANEOUS↗

ASC Co-design Proxy App Strategy

The NNSA labs representing the ASC program are coordinating on the development and release of a portfolio of proxy applications to be used in co-design. While each lab will develop their own set of proxy apps that represent the applications and algorithms of interest to them, we propose a common set of criteria to describe and measure proxy apps, and a strategy aimed at minimizing duplication and maximizing impact across the entire ASC proxy app portfolio. Proxy application (or proxy apps for short) is a catchall term for the simplification of characteristics of real applications that are of interest to DOE. These proxy apps are used in the co-design process as concrete examples for component and system designers to understand our software requirements. The proxy apps are generally openly available to the co-design community and, unlike benchmarks, are intended to facilitate the two-way communication required by co-design to optimally evaluate trade-offs in the system hardware, and inform the developer community of programming best practices for emerging architectures.

97 MATHEMATICS AND COMPUTING↗

Flexible Pilot Jobs Framework for Distributed High Throughput Computing

Experimental particle physics has been at the forefront of analyzing the world’s largest datasets for decades. The high-energy physics (HEP) community was among the first to develop suitable software and computing tools for this purpose. GlideinWMS is a Glidein-based workload management system whose purpose is to provide experiments like CMS at CERN, DUNE at Fermilab, and others, a way to access and efficiently use vast amounts of computing resources. This system wants to provide a simple way to submit jobs to a set of computing resources, that will be provided to users behind the scenes. Glideins are the pilot jobs executed on the worker nodes at the grid sites, performing operations such as hardware detection, environment setup, and error handling. After all these operations, they will launch the actual user job. Many grid sites are supported, such as shared clusters, Google CE, and AWS. My internship aimed to design and code a flexible pilot jobs framework that will replace the one used by GlideinWMS, developing a modular and flexible skeleton of the Glidein and adding further functionalities. My project also focused on the application of machine learning techniques as support to this management system.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Generic robot architecture

The present invention provides methods, computer readable media, and apparatuses for a generic robot architecture providing a framework that is easily portable to a variety of robot platforms and is configured to provide hardware abstractions, abstractions for generic robot attributes, environment abstractions, and robot behaviors. The generic robot architecture includes a hardware abstraction level and a robot abstraction level. The hardware abstraction level is configured for developing hardware abstractions that define, monitor, and control hardware modules available on a robot platform. The robot abstraction level is configured for defining robot attributes and provides a software framework for building robot behaviors from the robot attributes. Each of the robot attributes includes hardware information from at least one hardware abstraction. In addition, each robot attribute is configured to substantially isolate the robot behaviors from the at least one hardware abstraction.

Bruemmer, David J↗

Supporting multiple hardware architectures at CMS: the integration and validation of POWER9

Computing resources in the Worldwide LHC Computing Grid (WLCG) have been based entirely on the x86 architecture for more than two decades. In the near future, however, heterogeneous non-x86 resources, such as ARM, POWER and Risc-V, will become a substantial fraction of the resources that will be provided to the LHC experiments, due to their presence in existing and planned world-class HPC installations. The CMS experiment, one of the four large detectors at the LHC, has started to prepare for this situation, with the CMS software stack (CMSSW) already compiled for multiple architectures. In order to allow for a production use, the tools for workload management and job distribution need to be extended to be able to exploit heterogeneous architectures. Profiting from the opportunity to exploit the first sizable IBM Power9 allocation available on Marconi100 HPC system at CINECA, CMS developed all the needed modifications to the CMS workload management system. After a successful proof of concept, a full physics validation has been performed in order to bring the system in production. The experiences are of very high value, when it comes to commissioning of the similar (even larger) Summit HPC system at Oak Ridge, where CMS is also expecting a resource allocation. Moreover the compute power of those systems is being provided also via GPUs and this represents an extremely valuable opportunity to exploit the offloading capability already implemented in CMSSW. The status of the current integration including the exploitation of the GPUs, the results of the validation as well as the future plans will be shown and discussed.

Boccali, Tommaso [INFN, Pisa]↗

A Conceptual Framework for HPC Operational Data Analytics

This paper provides a broad framework for under- standing trends in Operational Data Analytics (ODA) for High- Performance Computing (HPC) facilities. The goal of ODA is to allow for the continuous monitoring, archiving, and analysis of near real-time performance data, providing immediately actionable information for multiple operational uses. In this work, we combine two models to provide a comprehensive HPC ODA framework: one is an evolutionary model of analytics capabilities that consists of four types, which are descriptive, diagnostic, predictive and prescriptive, while the other is a four- pillar model for energy-efficient HPC operations that covers facility, system hardware, system software, and applications. This new framework is then overlaid with a description of current development and production deployments of ODA within leading- edge HPC facilities. Finally, we perform a comprehensive survey of ODA works and classify them according to our framework, in order to demonstrate its effectiveness.

Netti, Alessio↗

WE-G-BRA-06: Application of Systems and Control Theory-Based Hazard Analysis to Radiotherapy

Purpose: The process of delivering radiation occurs in a complex socio-technical system heavily reliant on human operators. Furthermore, both humans and software are notoriously challenging to account for in traditional hazard analysis models. High reliability industries such as aviation have approached this problem through using hazard analysis techniques grounded in systems and control theory. The purpose of this work is to apply the Systems Theoretic Accident Model Processes (STAMP) hazard model to radiotherapy. In particular, the System-Theoretic Process Analysis (STPA) approach is used to perform a hazard analysis of a proposed on-line adaptive cranial radiosurgery procedure that omits the CT Simulation step and uses only CBCT for planning, localization, and treatment. Methods: The STPA procedure first requires the definition of high-level accidents and hazards leading to those accidents. From there, hierarchical control structures were created followed by the identification and description of control actions for each control structure. Utilizing these control structures, unsafe states of each control action were created. Scenarios contributing to unsafe control action states were then identified and translated into system requirements to constrain process behavior within safe boundaries. Results: Ten control structures were created for this new CBCT-only process which covered the areas of hospital and department management, treatment design and delivery, and vendor service. Twenty three control actions were identified that contributed to over 80 unsafe states of those control actions resulting in over 220 failure scenarios. Conclusion: The interaction of people, hardware, and software are highlighted through the STPA approach. STPA provides a hierarchical model for understanding the role of management decisions in impacting system safety so that a process design requirement can be traced back to the hazard and accident that it is intended to mitigate. Varian Medical Systems, Inc.

60 APPLIED LIFE SCIENCES↗

WBS STPR 04 Milestone 4 Report

This report documents the completion of milestone STPRO4-4 Kokkos back-ends research, collaborations, development, optimization, and documentation. The Kokkos team updated its existing backend to support the software stack and hardware of DOE's Sierra, Summit and Astra machines. They also collaborated with ECP PathForward vendors on developing backends for possible exa-scale architectures. Furthermore, the team ramped up its engagement with the ISO/C++ committee to accelerate the adoption of features important for the HPC community into the C++ standard.

97 MATHEMATICS AND COMPUTING↗

WBS STPR 04 Milestone 4 Report

This report documents the completion of milestone STPRO4-4 Kokkos back-ends research, collaborations, development, optimization, and documentation. The Kokkos team updated its existing backend to support the software stack and hardware of DOE's Sierra, Summit and Astra machines. They also collaborated with ECP PathForward vendors on developing backends for possible exa-scale architectures. Furthermore, the team ramped up its engagement with the ISO/C++ committee to accelerate the adoption of features important for the HPC community into the C++ standard.

97 MATHEMATICS AND COMPUTING↗

MO-PIS-Exhibit Hall-01: Imaging: CT Dose Optimization Technologies I

Partners in Solutions is an exciting new program in which AAPM partners with our vendors to present practical “hands-on” information about the equipment and software systems that we use in our clinics. The imaging topic this year is CT scanner dose optimization capabilities. Note that the sessions are being held in a special purpose room built on the Exhibit Hall Floor, to encourage further interaction with the vendors. Dose Optimization Capabilities of GE Computed Tomography Scanners Presentation Time: 11:15 – 11:45 AM GE Healthcare is dedicated to the delivery of high quality clinical images through the development of technologies, which optimize the application of ionizing radiation. In computed tomography, dose management solutions fall into four categories: employs projection data and statistical modeling to decrease noise in the reconstructed image - creating an opportunity for mA reduction in the acquisition of diagnostic images. Veo represents true Model Based Iterative Reconstruction (MBiR). Using high-level algorithms in tandem with advanced computing power, Veo enables lower pixel noise standard deviation and improved spatial resolution within a single image. Advanced Adaptive Image Filters allow for maintenance of spatial resolution while reducing image noise. Examples of adaptive image space filters include Neuro 3-D filters and Cardiac Noise Reduction Filters. AutomA adjusts mA along the z-axis and is the CT equivalent of auto exposure control in conventional x-ray systems. Dynamic Z-axis Tracking offers an additional opportunity for dose reduction in helical acquisitions while SmartTrack Z-axis Tracking serves to ensure beam, collimator and detector alignment during tube rotation. SmartmA provides angular mA modulation. ECG Helical Modulation reduces mA during the systolic phase of the heart cycle. SmartBeam optimization uses bowtie beam-shaping hardware and software to filter off-axis x-rays - minimizing dose and reducing x-ray scatter. The DICOM Radiation Dose Structured Report (RDSR) generates a dose report at the conclusion of every examination. Dose Check preemptively notifies CT operators when scan parameters exceed user-defined dose thresholds. DoseWatch is an information technology application providing vendor-agnostic dose tracking and analysis for CT (and all other diagnostic x-ray modalities) SnapShot Pulse improves coronary CTA dose management. VolumeShuttle uses two acquisitions to increase coverage, decrease dose, and conserve on contrast administration. Color-Coding for Kids applies the Broselow-Luten Pediatric System to facilitate pediatric emergency care and reduce medical errors. FeatherLight achieves dose optimization through pediatric procedure-based protocols. Adventure Series scanners provide a child-friendly imaging environment promoting patient cooperation with resultant reduction in retakes and patient motion. Philips CT Dose Optimization Tools and Advanced Reconstruction Presentation Time: 11:45 ‘ 12:15 PM The first part of the talk will cover “Dose Reduction and Dose Optimization Technologies” present in Philips CT Scanners. The main Technologies to be presented include: DoseRight and tube current modulation (DoseRight, Z-DOM, 3D-DOM, DoseRight Cardiac) Special acquisition modes Beam filtration and beam shapers Eclipse collimator and ClearRay collimator NanoPanel detector DoseRight will cover automatic tube current selection that automatically adjusts the dose for the individual patient. The presentation will explore the modulation techniques currently employed in Philips CT scanners and will include the algorithmic concepts as well as illustrative examples. Modulation and current selection technologies to be covered include the Automatic Current Selection component of DoseRight, ZDOM longitudinal dose modulation, 3D-DOM (combination of longitudinal and rotational dose modulation), Cardiac Dose right (an ECG based dose modulation scheme), and the DoseRight Index (DRI) IQ index. The special acquisition modes covers acquisition techniques such as prospective gating that is designed to reduce exposure to the patient through the Cardiac Step and Shoot scan mode. This mode can substitute the much higher dose retrospective scan modes for certain types of cardiac imaging. The beam filtration and beam shaper portion will discuss the variety of filtration and beam shaping configurations available on Philips scanners. This topic includes the x-ray beam characteristics, tube filtration as well as dose compensator characteristics. The Eclipse collimator, ClearRay collimator and the NanoPanel detector portion will discuss additional technologies specific to wide coverage CT that address some of the unique challenges encountered and techniques employed to optimize image quality and optimize dose utilization. The Eclipse collimator reduces extraneous exposure by actively blocking the radiation tails at either end of helical scans that do not contribute to the image generation. The ClearRay collimator and the NanoPanel detector optimize the quality of the signal that reaches the detectors by addressing the increased scattered radiation present in wide coverage and the NanoPanel detector adds superior electronic noise characteristics valuable when imaging at a low dose level. The second part of the talk will present “Advanced Reconstruction Technologies” currently available on Philips CT Scanners. The talk will cover filtered back projection (FBP), iDose4 and Iterative Model Reconstruction (IMR). Each reconstruction method will include a discussion of the algorithm as well as similarities and differences between the algorithms. Examples illustrating the merits of each algorithm will be presented, and techniques and metrics to characterize the performance of each type of algorithm will be presented. The Filtered Back projection portion will discuss and provide a brief summary of relevant standard image reconstruction techniques in common use, and discuss the common tradeoffs when using the FBP algorithm. The iDose4 portion will present the algorithms used for iDose4 as well the different levels. The meaning of different levels of iDose4 available will be presented and quantified. Guidelines for selection iDose4 parameters based on the imaging need will be explained. The different image quality goals available with iDose4 and specifically how iDose4 enables noise reduction, spatial resolution improvement or both will be explained. The approaches to leveraging the benefits of iDose4 such as improved spatial resolution, decreased noise, and artifact prevention will be described and quantified; and measurements and metrics behind the improvements will be presented. The image quality benefits in specific imaging situations as well as how to best combine the technology with other dose reduction strategies to ensure the best image quality at a given dose level will be presented. Insight into the IMR algorithm as well as contrast to the iDose4 techniques and performance characteristics will be discussed. Metrics and techniques for characterizing this class of algorithm and IQ performance will be presented. The image quality benefits and the dose reduction capabilities of IMR will be explored. Illustrative examples of the noise reduction, spatial resolution improvement, and low contrast detectability improvements of the reconstruction method will be presented: clinical cases and phantom measurements demonstrating the benefits of IMR in the areas of low dose imaging, spatial resolution and low contrast resolution are discussed and the technical details behind the measurements will be presented compared to both iDose4 and traditional filtered back projection (FBP)

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

A Toolkit for ARB to Integrate Custom Databases and Externally Built Phylogenies

Researchers are perpetually amassing biological sequence data. The computational approaches employed by ecologists for organizing this data (e.g. alignment, phylogeny, etc.) typically scale nonlinearly in execution time with the size of the dataset. This often serves as a bottleneck for processing experimental data since many molecular studies are characterized by massive datasets. To keep up with experimental data demands, ecologists are forced to choose between continually upgrading expensive in-house computer hardware or outsourcing the most demanding computations to the cloud. Outsourcing is attractive since it is the least expensive option, but does not necessarily allow direct user interaction with the data for exploratory analysis. Desktop analytical tools such as ARB are indispensable for this purpose, but they do not necessarily offer a convenient solution for the coordination and integration of datasets between local and outsourced destinations. Therefore, researchers are currently left with an undesirable tradeoff between computational throughput and analytical capability. To mitigate this tradeoff we introduce a software package to leverage the utility of the interactive exploratory tools offered by ARB with the computational throughput of cloud-based resources. Our pipeline serves as middleware between the desktop and the cloud allowing researchers to form local custom databases containing sequences and metadata from multiple resources and a method for linking data outsourced for computation back to the local database. Furthermore, a tutorial implementation of the toolkit is provided in the supporting information, S1 Tutorial.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

An agent-based hardware-in-the-loop simulation framework for building controls

In general, a hardware-in-the-loop (HIL) building simulation has lower cost and fewer practical limitations (e.g., scheduling issues) than field tests in occupied buildings, while also overcoming limitations of simulations alone by capturing the full behavior of some physical systems, equipment, and components. However, the implementation of an HIL can be difficult due to the scarcity of appropriate tools. Here, this paper presents an agent-based framework for HIL simulation. It can be used for investigation of controller performance via controller-in-the-loop simulations and also HIL for system synthesis. In the latter case, both controllers and major equipment participate in tests to ensure that dynamics of equipment operation are correctly captured in addition to controller performance. The HIL simulation framework presented allows such actual physical parts to be included in the framework while representing others for which behaviors are better known and modeled in simulation models. The mechanism implemented in the framework to synchronize simulations in software with real-time operation of physical equipment is described. Lastly, as an example, use of the HIL simulation framework is illustrated through a brief study of speed control of the supply fan in the air handling unit of a variable-air-volume building heating, ventilating and air-conditioning system.

42 ENGINEERING↗

Obstacles to Practical Digital Supply Chain Risk Management in the Energy Sector

Cyber supply chain risk management (C-SCRM) programs must consider operations that depend on the lifecycles of digital components such as hardware, firmware, software, and services. We integrate academic literature, historical incidents, and existing standards to identify obstacles faced by C-SCRM programs.

Business Process Management & Integration↗

Even Higher-Level Synthesis: An Exploration of AI Hardware Accelerators using HLS4ML

With the rise of artificial intelligence, the popularization of deep learning, and a constantly evolving industry, the demand for flexible and efficient tools has never been greater. As algorithms grow more complex, their runtime and energy consumption increase exponentially. Customized hardware accelerators, long used for specific mathematical operations, remain essential for managing modern applications' computational and power demands. Hardware accelerators can speed up complex computations by orders of magnitude, but their manual design and verification processes are often challenging and time-consuming. High-Level Synthesis (HLS) provides a solution by transforming high-level algorithm descriptions, typically written in C++ or SystemC, into synthesizable RTL suitable for hardware implementation. This approach reduces development time for RTL engineers while offering flexibility beyond what traditional handwritten RTL can provide. We extended this capability to the machine-learning domain with the open-source framework hls4ml, which allows neural networks trained in Python frameworks like Tensorflow or PyTorch to be synthesized into efficient hardware representations for the traditional FPGA and ASIC flows. This breakthrough addresses the growing need for reduced design turnaround and easy verification of ML hardware accelerators with low latency and power efficiency constraints. During this tutorial, we will demonstrate how Python complements HLS by simplifying the ML design process, bridging the gap between software and hardware development. Attendees will explore how we translate neural networks modeled in Python into fixed-point C++ models suitable for HLS workflows. We will dive into strategies like Value-Range Analysis and Quantization-Aware Training, which optimize these designs for deployment and evaluate their accuracy, power consumption, and energy efficiency. To exemplify these concepts, experts from Fermilab will share their experiences applying this technology to high-energy physics experiments, where real-time, low-latency processing is critical. Over the years, Fermilab engineers have demonstrated how deep neural networks, optimized for hardware using hls4ml, can meet the stringent requirements of trigger systems at the CERN Large Hadron Collider. These systems rely on rapid decision-making to process immense data volumes while retaining only the most relevant events for further analysis. The application of hls4ml has also been extended to innovative technologies like smart pixel arrays. These smart pixels integrate ML inference capabilities directly into sensor devices, enabling localized data processing at the pixel level. This approach drastically reduces the need to transmit raw data to external processing units, significantly decreasing power consumption and latency. By embedding neural networks within the pixel architecture, the smart pixels can identify and prioritize relevant data in real time, providing a highly efficient solution for edge computing in scenarios such as particle detectors and imaging systems. Fermilab's work highlights the potential of hardware-accelerated ML in scenarios where both speed and power efficiency are mission-critical. Through this tutorial, attendees will gain valuable insights into the challenges and solutions of deploying ML in hardware. Understanding how HLS and hls4ml streamline the development of neural network-based hardware accelerators is fundamental for the industry's future. Participants will learn how these technologies are shaping the future of AI and scientific computing.

Di Guglielmo, Giuseppe [Fermilab]↗

Hierarchical Model-Free Transactive Control of Building Loads to Support Grid Services

Residential buildings consume 4.4 quads of electricity annually, approximately 37% of the total electricity consumption in the United States. This represents a vast resource that can be used for demand management and other ancillary services. This project aims to develop a robust, scalable hierarchical transactional control mechanism incorporating elements of model-free control (MFC) and game theory to harness buildings to provide ancillary services to the grid. This approach is being taken to address the challenges of incorporating traditional transactional control schemes into existing buildings. The challenges include small individual building sizes requiring aggregation of many buildings, unpredictable energy usage that makes model identification difficult, and satisfying the sensitive occupant comfort constraints. In the proposed approach, by separating the control mechanism into two layers above and below the load aggregator, MFC can be used below the aggregator to modulate flexible building loads in response to pricing signals with guaranteed performance. This allows the burden of identifying an accurate model of the system to be shifted to the above-aggregator layer, where fluctuations in individual building usage have less impact on predicted building system behavior. Game theory concepts can then be used to determine pricing curves and control signals among regional aggregators. Managing this control in a game-theoretic approach will allow us to build in financial incentives that increase customer engagement. Additionally, the usage of MFC necessitates less burdensome computational and communication requirements, thus, it is easily deployable on small, embedded devices. In a broader sense, developing a strategy capable of effectively incorporating residential and small commercial buildings will allow greater throughput of existing and emerging grid services in addition to future transactive energy grid management methods. Using MFC within a hierarchical control architecture will allow the shifting of existing forecasting challenges to an aggregate level, where dynamics are slower and more predictable. This will enable a smooth interface between the grid services requests of utilities and the reliable control required by participating buildings. MFC, which supports distributed control architecture, permits a scalable solution that can be deployed to neighborhood-size systems as well as individual buildings. This project focuses on three objectives: (1) developing the mathematical framework, algorithm toolkit, and software toolset of the two-layer transactive control testbed; (2) developing a scalable solution for application over many residential and small-size commercial buildings with sparse distributed communication; and (3) field testing and implementation on hardware of the control strategies developed in the previous two objectives. The research and development activities are focused and designed to be impactful within the relevant 2025 targets timeframe. An open-source control framework for exploiting variability and dispatchability of building loads will be delivered as the outcome of the project. This capability enables greater participation of loads in electricity markets and ancillary services that are both useful for the utility and financially beneficial for building owners.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Advanced Models and Fast Simulation Algorithms for MARS Plants

With the successful commercialization of modular multilevel converters (MMC) and rapid increase in photovoltaic (PV) penetration and energy storage systems (ESS), one of the research areas being explored dynamically is integration of PV systems and ESS systems to MMCs. On this front, an integrated development of PV plants, ESS, and alternating- and direct-current (ac/dc) systems through Multi-Port Autonomous Reconfigurable Solar (MARS) plant is proposed and being researched upon. However, simulation of MARS systems on traditional existing software takes very long time (order of days) due to the presence of a large number of states and non-linear devices. Therefore, development of fast high fidelity models is of paramount importance to test the hardware design, control design and planning of power system expansions. In this context, two research objectives are addressed in this paper: firstly, an ultra fast single- or multi- CPU simulation algorithm to simulate the MARS system based on state-space models, hybrid discretization algorithm with a relaxation technique that reduces the imposed computational burden and a multi-rate method is presented. Secondly, a reduced order model of MARS is also presented. Both the developed techniques are validated with respect to reference PSCAD/EMTDC model.

Marthi, Phani Ratna Vanamali↗

Resilience Design Patterns: A Structured Approach to Resilience at Extreme Scale

Reliability is a serious concern for future extreme-scale high-performance computing (HPC) systems. Projections based on the current generation of HPC systems and technology roadmaps suggest the prevalence of very high fault rates in future systems. While the HPC community has developed various resilience solutions, application-level techniques as well as system-based solutions, the solution space remains fragmented. There are no formal methods and metrics to integrate the various HPC resilience techniques into composite solutions, nor are there methods to holistically evaluate the adequacy and efficacy of such solutions in terms of their protection coverage, and their performance \& power efficiency characteristics. Additionally, few of the current approaches are portable to newer architectures and software environments that will be deployed on future systems. In this paper, we develop a structured approach to the design, evaluation and optimization of HPC resilience using the concept of design patterns. A design pattern is a general repeatable solution to a commonly occurring problem. We identify the problems caused by various types of faults, errors and failures in HPC systems and the techniques used to deal with these events. Each well-known solution that addresses a specific HPC resilience challenge is described in the form of a pattern. We develop a complete catalog of such resilience design patterns, which may be used by system architects, system software and tools developers, application programmers, as well as users and operators as essential building blocks when designing and deploying resilience solutions. We also develop a design framework that enhances a designer's understanding the opportunities for integrating multiple patterns across layers of the system stack and the important constraints during implementation of the individual patterns. It is also useful for defining mechanisms and interfaces to coordinate flexible fault management across hardware and software components. The resilience patterns and the design framework also enable exploration and evaluation of design alternatives and support optimization of the cost-benefit trade-offs among performance, protection coverage, and power consumption of resilience solutions. Here, the overall goal of this work is to establish a systematic methodology for the design and evaluation of resilience technologies in extreme-scale HPC systems that keep scientific applications running to a correct solution in a timely and cost-efficient manner despite frequent faults, errors, and failures of various types.

97 MATHEMATICS AND COMPUTING↗

5G Enabled Energy Innovation: Advanced Wireless Networks for Science (Workshop Report)

Rapidly expanding, new telecommunications infrastructure based on 5G technologies will disrupt and transform how we design, build, operate, and optimize scientific infrastructure and the experiments and services enabled by that infrastructure, from continental-scale sensor networks to centralized scientific user facilities, from intelligent Internet of Things devices to supercomputers. Concurrently, 5G will introduce, or exacerbate, challenges related to protecting infrastructure and associated scientific data as well as to fully leveraging opportunities related to expanded infrastructure scale and complexity. The U.S. Department of Energy (DOE) Office of Science operates scientific infrastructure, supporting some of the nation’s most advanced intellectual discoveries, spanning the country and including 30 world-class user facilities from supercomputers to accelerators. Along with field experiments and remote observatories, every aspect of DOE’s scientific enterprise will be affected by 5G, which amounts to a complete renovation of the underpinnings of the nation’s information infrastructure. In this report we explore the scientific opportunities and new research challenges associated with 5G, ranging from scalability to heterogeneity to cybersecurity. The rapid commercial deployment of 5G opens the opportunity to rethink and reinvent DOE’s scientific infrastructure and experimentation, from intelligent sensor networks at unprecedented scales to a digital continuum of cyberinfrastructure spanning low-power sensors, high-performance computing embedded within and at the edge of the network, and DOE’s large-scale user instrument and computing facilities. New programming paradigms, workflow and data frameworks, and AI-based system design, operation, and autonomous adaptation and optimization will be necessary in order to exploit these new opportunities. Field deployments and centralized scientific instruments can also be revolutionized, moving (without traditional performance penalties) from wired to wireless connectivity for data and control systems, improving flexibility, and opening new sensing modalities, including the use of the 5G electromagnetic spectrum itself as an environmental probe. For DOE science, in contrast to commercial 5G applications and settings, devices will be deployed in extreme environments such as cryogenically cooled instrument control systems and in remote settings with harsh conditions, requiring the design of new materials for RF communication and edge processing to operate in these regimes. Concurrently, 5G infrastructure comprises both hardware and sophisticated software systems - currently closed and proprietary. The cybersecurity challenges to 5G-empowered reinvention mirror the complexity and variety of new 5G features, from virtualization to private network slices to ubiquitous access. Research is also needed in order to accelerate the development of secure and open 5G software infrastructure, reducing reliance on hardware and software produced outside the United States and providing the transparency and rigorous evaluation and testing afforded through open software. Twelve broad research thrusts are laid out in four chapters, with a companion fifth chapter (and three additional research thrusts) underscoring the needs and opportunities for an aggressive testbed program co-designed by networking experts and scientists involved in the 15 research thrusts. The urgency of undertaking this research is fueled by a global, accelerating deployment of new telecommunications infrastructure that is designed for entertainment and commercial applications - barely scratching the surface of what 5G can do to extend U.S. leadership in scientific discovery.

42 ENGINEERING↗