SEARCH · Engineering Papers
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.
This paper describes the design of Chapel, the Cascade High Productivity Language, which is being developed in the DARPA-funded HPCS project Cascade led by Cray Inc. Chapel pushes the state-of-the-art in languages for HEC system programming by focusing on productivity, in particular by combining the goal of highest possible object code performance with that of programmability offered by a high-level user interface.
This compiler for ARTS (CARTS) is framework designed to connect a high productive languages with a distributed fine grained runtime that run effectively across clusters. It is built using MLIR and the LLVM infrastructure and it can be used as a platform to test static analysis ideas mapped towards distributed environments with novel technologies
This project developed the LAPIS compiler framework, built on the Multilevel Intermediate Representation (MLIR), to optimize sparse linear algebra operations and support performance portability across diverse architectures. The main innovation of LAPIS is the Kokkos dialect, which allows for lowering codes from a high productivity language to different architectures in an elegant way. The dialect also allows the conversion of lower-level MLIR code to C++ Kokkos code, facilitating the integration of scientific machine learning (SciML) models into applications. To extend LAPIS for distributed memory architectures, a new partition dialect was created to manage the distribution of sparse tensors and express communication patterns for sparse linear algebra operations. This dialect also supports the distributed execution of operators and includes algorithmic optimizations to minimize communication to improve performance. The project also demonstrates that MLIR can enable effective linear algebra-level optimizations, improving performance on different GPUs for both sparse and dense linear algebra kernels. Key applications of LAPIS include sparse linear algebra and graph kernels, TenSQL, a relational database management solution built on GraphBLAS, and the development of subgraph isomorphism and monomorphism kernels, showcasing performance portability. In summary, the LAPIS framework supports productivity, performance, portability, and distributed memory execution, while also enabling linear algebra-level optimizations that are challenging in traditional programming languages, with successful applications ranging from simple sparse linear algebra to complex graph kernels.
This paper presents a reusable design of a data distribution framework for data parallel high performance applications. We are implementing the design in the context of the Chapel high productivity programming language. Distributions in Chapel are a means to express locality in systems composed of large numbers of processor and memory components connected by a network. Since distributions have a great effect on,the performance of applications, it is important that the distribution strategy can be chosen by a user. At the same time, high productivity concerns require that the user is shielded from error-prone, tedious details such as communication and synchronization. We propose an approach to distributions that enables the user to refine a language-provided distribution type and adjust it to optimize the performance of the application. Additionally, we conceal from the user low-level communication and synchronization details to increase productivity. To emphasize the generality of our distribution machinery, we present its abstract design in the form of a design pattern, which is independent of a concrete implementation. To illustrate the applicability of our distribution framework design, we outline the implementation of data distributions in terms of the Chapel language.
Distributed memory architectures offer high levels of performance and flexibility, but have proven awkard to program. Current languages for nonshared memory architectures provide a relatively low level programming environment, and are poorly suited to modular programming, and to the construction of libraries. A set of language primitives designed to allow the specification of parallel numerical algorithms at a higher level is described. Tensor product array computations are focused on along with a simple but important class of numerical algorithms. The problem of programming 1-D kernal routines is focused on first, such as parallel tridiagonal solvers, and then how such parallel kernels can be combined to form parallel tensor product algorithms is examined.
In order to increase the efficiency of process control data processing, it is necessary to enhance the productivity of real time high level languages and to automate the task administration, because presently 60 percent or more of the applications are still programmed in assembly languages. This may be achieved by migrating apt functions for the support of process control oriented languages into the hardware, i.e., by new architectures. Whereas numerous high level languages have already been defined or realized, there are no investigations yet on hardware assisted implementation of real time features. The requirements to be fulfilled by languages and operating systems in hard real time environment are summarized. A comparison of the most prominent languages, viz. Ada, HAL/S, LTR, Pearl, as well as the real time extensions of FORTRAN and PL/1, reveals how existing languages meet these demands and which features still need to be incorporated to enable the development of reliable software with predictable program behavior, thus making it possible to carry out a technical safety approval. Accordingly, Pearl proved to be the closest match to the mentioned requirements.
Production LLM serving must simultaneously deliver high throughput, low latency, and sufficient context capacity under non-stationary traffic and mixed request requirements. Data parallelism (DP) maximizes throughput by running independent replicas, while tensor parallelism (TP) reduces per-request latency and pools memory for long-context inference. However, existing serving stacks typically commit to a static parallelism configuration at deployment; adapting to bursts, priorities, or long-context requests is often disruptive and slow. We present Flying Serving, a vLLM-based system that enables online DP-TP switching without restarting engine workers. Flying Serving makes reconfiguration practical by virtualizing the state that would otherwise force data movement: (i) a zero-copy Model Weights Manager that exposes TP shard views on demand, (ii) a KV Cache Adaptor that preserves request KV state across DP/TP layouts, (iii) an eagerly initialized Communicator Pool to amortize collective setup, and (iv) a deadlock-free scheduler that coordinates safe transitions under execution skew. Across three popular LLMs and realistic serving scenarios, Flying Serving improves performance by up to 4.79 × under high load and 3.47 × under low load while supporting latency- and memory-driven requests.
Expert systems are computer programs which emulate human expertise in well defined problem domains. The potential payoff from expert systems is high: valuable expertise can be captured and preserved, repetitive and/or mundane tasks requiring human expertise can be automated, and uniformity can be applied in decision making processes. The C Language Integrated Production System (CLIPS) is an expert system building tool, developed at the Johnson Space Center, which provides a complete environment for the development and delivery of rule and/or object based expert systems. CLIPS was specifically designed to provide a low cost option for developing and deploying expert system applications across a wide range of hardware platforms. The commercial potential of CLIPS is vast. Currently, CLIPS is being used by over 5,000 individuals throughout the public and private sector. Because the CLIPS source code is readily available, numerous groups have used CLIPS as the basis for their own expert system tools. To date, three commercially available tools have been derived from CLIPS. In general, the development of CLIPS has helped to improve the ability to deliver expert system technology throughout the public and private sectors for a wide range of applications and diverse computing environments.
SOCIAL is an advanced, object-oriented development tool for integrating intelligent and conventional applications across heterogeneous hardware and software platforms. SOCIAL defines a family of 'wrapper' objects called agents, which incorporate predefined capabilities for distributed communication and control. Developers embed applications within agents and establish interactions between distributed agents via non-intrusive message-based interfaces. This paper describes a predefined SOCIAL agent that is specialized for integrating C Language Integrated Production System (CLIPS)-based applications. The agent's high-level Application Programming Interface supports bidirectional flow of data, knowledge, and commands to other agents, enabling CLIPS applications to initiate interactions autonomously, and respond to requests and results from heterogeneous remote systems. The design and operation of CLIPS agents are illustrated with two distributed applications that integrate CLIPS-based expert systems with other intelligent systems for isolating and mapping problems in the Space Shuttle Launch Processing System at the NASA Kennedy Space Center.
ChatHPC democratizes large language models for the high-performance computing (HPC) community by providing the infrastructure, ecosystem, and knowledge needed to apply modern generative AI technologies to rapidly create specific capabilities for critical HPC components while using relatively modest computational resources. Our divide-and-conquer approach focuses on creating a collection of reliable, highly specialized, and optimized AI assistants for HPC based on the cost-effective and fast Code Llama fine-tuning processes and expert supervision. We target major components of the HPC software stack, including programming models, runtimes, I/O, tooling, and math libraries. Thanks to AI, ChatHPC provides a more productive HPC ecosystem by boosting important tasks related to portability, parallelization, optimization, scalability, and instrumentation, among others. With relatively small datasets (on the order of KB), the AI assistants, which are created in a few minutes by using one node with two NVIDIA H100 GPUs and the ChatHPC library, can create new capabilities with Meta’s 7-billion parameter Code Llama base model to produce high-quality software with a level of trustworthiness of up to 90% higher than the 1.8-trillion parameter OpenAI ChatGPT-4o model for critical programming tasks in the HPC software stack.
PIFLEX is a pipelined-image processor that can perform elaborate computations whose exact nature is not fixed in the hardware, and that can handle multiple images. A wire-wrapped prototype PIFEX module has been produced and debugged, using a version of the convolver composed of three custom VLSI chips (plus the line buffers). A printed circuit layout is being designed for use with a single-chip convolver, leading to production of a PIFEX with about 120 modules. A high-level language for programming PIFEX has been designed, and a compiler will be written for it. The camera calibration software has been completed and tested. Two more terms in the camera model, for lens distortion, probably will be added later. The acquisition and tracking system has been designed and most of it has been coded in Pascal for the MicroVAX-II. The feature tracker, motion stereo module and stereo matcher have executed successfully. The model matcher is still under development, and coding has begun on the tracking initializer. The object tracker was running on a different computer from the VAX, and preliminary runs on real images have been performed there. Once all modules are working, optimization and integration will begin. Finally, when a sufficiently large PIFEX is available, appropriate parts of acquisition and tracking, including much of the feature tracker, will be programmed into PIFEX, thus increasing the speed and robustness of the system.
Vulnerability and patch management is an integral part of a robust cybersecurity program, yet it grows increasingly complex due to the sheer amount of data that must be analyzed. Particularly in Operational Technology (OT) environments, analysis must be done manually because of the lack of automated solutions. Additionally, there are many steps in this process, from the initial discovery of the vulnerability to the implementation of its remediation, and each step in the process requires different data in order to be performed effectively. In this work, we provide approaches and strategies to assist operators in industrial or OT environments throughout the vulnerability management cycle. Security advisories provide key information about mitigation strategies, or actions that can be taken when a patch is unavailable or cannot be installed. Details of these strategies are not shared in public vulnerability databases and must be found manually. We approach this problem by designing a solution to automatically identify that information within vendor security advisories and retrieve it for operator use. We start with an approach that requires domain-specific knowledge of certain frequently-seen reference websites. Next, an approach that can work on an arbitrary website but relies on certain keywords. Finally, an approach that uses Natural Language Processing (NLP) methods and does not require specific knowledge or keywords. Each of these approaches is more general than its predecessor; we demonstrate high accuracy for all approaches Advisories also often contain details of affected products in non-standard or natural language formats. While this information can be easily understood when read by an operator, the non-standard format acts as a barrier to effective automation. We provide an approach for the first step in this process: identifying vendors in security advisories and mapping them to a standard framework for representing digital assets and software products. We evaluate five established string similarity algorithms, plus one of our own design that combines string similarity and information theory, on the task of mapping vendors to their corresponding entries in the Common Platform Enumeration (CPE) repository. Our results show that our proposed metric outperforms all others. Due to the constraints on time, finances, and personnel for organizations, Large Language Models (LLMs) may seem like attractive opportunities for security operators to speed up information gathering; however, it is still not clear whether LLMs can handle vulnerability management tasks well. To answer this question, we perform an empirical study of LLMs’ ability to provide consistent, accurate information about vulnerabilities in order to guide organizations in their adoption of LLMs. We observe poor performance for all models tested, suggesting that these models are not well-suited to the consistent retrieval of accurate vulnerability information. Finally, once vulnerabilities have been identified and any additional information has been obtained, operators must decide which remediation actions to implement based on their available resources. This already-complex problem becomes even more so when we consider that a vulnerability may have multiple avenues for remediation. We formulate this scenario as two knapsack problems and provide solutions, which we then compare against several existing strategies for vulnerability prioritization seen in real operational environments.
The Naval Postgraduate School has been conducting research into the design and testing of an Autonomous Underwater Vehicle (AUV). One facet of this research is to incrementally design a software architecture and implement it in an advanced testbed, the AUV II. As part of the high level architecture, a Mission Executor is being constructed using CLIPS (C Language Integrated Production System) version 5.0. The Mission Executor is an expert system designed to oversee progress from the AUV launch point to a goal area and back to the origin. It is expected that the executor will make informed decisions about the mission, taking into account the navigational path, the vehicle subsystem health, and the sea environment, as well as the specific mission profile which is downloaded from an offboard mission planner. Heuristics for maneuvering, avoidance of uncharted obstacles, waypoint navigation, and reaction to emergencies (essentially the expert knowledge of a submarine captain) are required. Many of the vehicle subsystems are modeled as objects using the CLIPS Object Oriented Language (COOL) embedded in CLIPS 5.0. Also, truth maintenance is applied to the knowledge base to keep configurations updated.
Over the last few years we have seen an increasing number of applications of Fuzzy Logic Controllers. These applications range from the development of auto-focus cameras, to the control of subway trains, cranes, automobile subsystems (automatic transmissions), domestic appliances, and various consumer electronic products. In summary, we consider a Fuzzy Logic Controller to be a high level language with its local semantics, interpreter, and compiler, which enables us to quickly synthesize non-linear controllers for dynamic systems.
The current generation of languages for the construction of knowledge-based systems as being at too low a level of abstraction is criticized, and the need for higher level languages for building problem solving systems is advanced. A notion of generic information processing tasks in knowledge-based problem solving is introduced. A toolset which can be used to build expert systems in a way that enhances intelligibility and productivity in knowledge acquistion and system construction is described. The power of these ideas is illustrated by paying special attention to a high level language called DSPL. A description is given of how it was used in the construction of a system called MPA, which assists with planning in the domain of offensive counter air missions.
This paper outlines a new telescope operations model that is intended to achieve low operating costs with high operating efficiency and high scientific productivity. The model is based on the existing Principal Astronomer approach used in conjunction with ATIS, a language for commanding remotely located automatic telescopes. This paper introduces the notion of an Associate Principal Astronomer, or APA. At the heart of the APA is automatic observation loading and scheduling software, and it is this software that is expected to help achieve efficient and productive telescope operations. The purpose of the APA system is to make it possible for astronomers to submit observation requests to and obtain resulting data from remote automatic telescopes, via the Internet, in a highly-automated way that minimizes human interaction with the system and maximizes the scientific return from observing time.
Explore the source record for details and available documents.
In order to implement high level languages whenever possible, a translator writing system of advanced design was developed. It is intended for routine production use by many programmers working on different projects. As well as a fairly conventional parser generator, it includes a system for the rapid generation of table driven code generators. The parser generator was developed from a prototype version. The translator writing system includes various tools for the management of the source text of a compiler under construction. In addition, it supplies various default source code sections so that its output is always compilable and executable. The system thereby encourages iterative enhancement as a development methodology by ensuring an executable program from the earliest stages of a compiler development project. The translator writing system includes PASCAL/48 compiler, three assemblers, and two compilers for a subset of HAL/S.