Results for “shared memory” — Page 27

DOE OSTI2024

P38 heterogeneous multi-tiled system with support for message queues (MoSAIC) v0.1

The proposed system is written in the hardware description language (HDL) verilog targeting an FPGA board. It is intended as a testbed to explore architecture tradeoffs in multi-tiled heterogeneous architectures. Although we target FPGAs, the system can be implemented as a monolithic SoC or a package comprised of many chiplets that are interconnected in the same package using a NoC. The proposed NoC is lightweight and follows an axi-lite interface. The endpoints of the NoC are a heterogeneous mix of "tiles" as endpoints that are general purpose processors, fixed function accelerators, and programmable accelerators. We assume that the network interfaces for the NoC endpoints are all addressable in a global name-space in that they represent an address range (for memory addresses) or a range of unique identifiers that are associated with each individual tile. This makes the functionality abstract from the standpoint of the NoC design details. Message queues offer a direct inter-processor interface between peer general purpose cores and diverse accelerators that comprise an SoC. Although they share the same NoC infrastructure for inter-tile communication within an SoC or SiP, the hardware message queues bypass the memory hierarchy and thus do not pollute the memory state or invoke the cache coherence mechanism.

Gonzalez, LouisaPatricia↗

DOE OSTI2021

Distributed directory of named data elements in coordination namespace

An approach is described that provides a distributed directory structure within a storage of an information handling system (a local node). A request is received with the request corresponding to a shared virtual address. The shared virtual address that is shared amongst a number of nodes that includes the local node and some remote nodes. A Global Address Space Directory (GASD) is retrieved that corresponds to a global virtual address space. The GASD is stored in a Coordination Namespace that is stored in a memory that is distributed amongst the nodes. A mapping that is included in the GASD is used to determine the node where the shared virtual address currently resides. The shared virtual address is then accessed from the node where it currently resides.

Johns, Charles Ray↗

DOE OSTI2024

Privacy-Preserving Control of Partitioned Energy Resources

Distributed energy resources are an increasingly important part of the electric grid. We examine the problem of partitioning a distributed energy resource among many users while providing privacy to them. In this model, clients can send requests to a server, the server can verify that the requests are valid and aggregate them, but it cannot see the actual values in the requests. Without privacy, each user is forced to reveal their daily schedule or energy use. Energy resources add a novel challenge that prior systems do not address: they require verifying limits on private power (a rate over time) and energy (a sum) values. Furthermore, the cryptographic mechanisms must run on embedded energy control systems. We describe Weft, a novel cryptographic system that verifies both power (rate) and energy (integral) constraints on private client values and aggregates them. The key insight behind the approach is to rely on additively homomorphic secret shares, which allows servers to compute sums from rates. We present 3 cryptographic proof systems with different system trade-off for embedded systems: bit-splitting proofs minimize memory use, sorting proofs minimize computation, and commitment proofs minimize network communication. Using bit-splitting proofs, it takes an IoT client using a CortexM microcontroller 4 minutes of compute time to privately control its share of an energy resource for a day at 20s granularity.

Laufer, Evan↗

NASA NTRS1989

Knowledge acquisition from natural language for expert systems based on classification problem-solving methods

It is shown how certain kinds of domain independent expert systems based on classification problem-solving methods can be constructed directly from natural language descriptions by a human expert. The expert knowledge is not translated into production rules. Rather, it is mapped into conceptual structures which are integrated into long-term memory (LTM). The resulting system is one in which problem-solving, retrieval and memory organization are integrated processes. In other words, the same algorithm and knowledge representation structures are shared by these processes. As a result of this, the system can answer questions, solve problems or reorganize LTM.

Gomez, Fernando↗

DOE OSTI2020

The parallelism motifs of genomic data analysis

Genomic datasets are growing dramatically as the cost of sequencing continues to decline and small sequencing devices become available. Enormous community databases store and share these data with the research community, but some of these genomic data analysis problems require large-scale computational platforms to meet both the memory and computational requirements. These applications differ from scientific simulations that dominate the workload on high-end parallel systems today and place different requirements on programming support, software libraries and parallel architectural design. For example, they involve irregular communication patterns such as asynchronous updates to shared data structures. We consider several problems in high-performance genomics analysis, including alignment, profiling, clustering and assembly for both single genomes and metagenomes. We identify some of the common computational patterns or ‘motifs’ that help inform parallelization strategies and compare our motifs to some of the established lists, arguing that at least two key patterns, sorting and hashing, are missing. This article is part of a discussion meeting issue ‘Numerical algorithms for high-performance computational science’.

59 BASIC BIOLOGICAL SCIENCES↗

NASA NTRS1995

Exploring the use of I/O nodes for computation in a MIMD multiprocessor

As parallel systems move into the production scientific-computing world, the emphasis will be on cost-effective solutions that provide high throughput for a mix of applications. Cost effective solutions demand that a system make effective use of all of its resources. Many MIMD multiprocessors today, however, distinguish between 'compute' and 'I/O' nodes, the latter having attached disks and being dedicated to running the file-system server. This static division of responsibilities simplifies system management but does not necessarily lead to the best performance in workloads that need a different balance of computation and I/O. Of course, computational processes sharing a node with a file-system service may receive less CPU time, network bandwidth, and memory bandwidth than they would on a computation-only node. In this paper we begin to examine this issue experimentally. We found that high performance I/O does not necessarily require substantial CPU time, leaving plenty of time for application computation. There were some complex file-system requests, however, which left little CPU time available to the application. (The impact on network and memory bandwidth still needs to be determined.) For applications (or users) that cannot tolerate an occasional interruption, we recommend that they continue to use only compute nodes. For tolerant applications needing more cycles than those provided by the compute nodes, we recommend that they take full advantage of both compute and I/O nodes for computation, and that operating systems should make this possible.

Kotz, David↗

NASA NTRS2012

Design for the Structure and the Mechanics of Moballs

The moball is envisioned to be a round, self-powered, and wind-driven multifunctioning sensor used in the Gone with the Wind ON-Mars (GOWON) [http://www.lpi.usra.edu/meetings/ marsconcepts2012/pdf/4238.pdf]: A Wind-Driven Networked System of Mobile sensors on Mars. The moballs would have sensing, processing, and communication capabilities. The moballs would perform in situ detection of key environmental elements such as vaporized water, trace gases, wind, dust, clouds, light and UV exposure, temperature, as well as minerals of interest, possible biosignatures, surface magnetic and electric fields, etc. The embedded various low-power micro instruments could include a Multispectral Microscopic Imager (to detect various minerals), a compact curved focal plane array camera (UV/Vis/NIR) with a large field of view, a compact UV/Visible spectrometer, a micro-weather station, etc. The moballs could communicate with each other and an orbiter. Their wind- or gravity-driven rolling movement could be used to harvest and store electric energy. They could also generate and store energy using the sunlight, when available, and the diurnal temperature variations on Mars. The moballs would be self-aware of their (and their neighbors ) positions, energy storage, and memory availability; they would have processing power and could intelligently cooperate with neighboring moballs by distributing tasks, sharing data, and fusing information. The major advantages of using the wind-driven and spherical moball network over rovers or other fixed sensor webs to explore Mars would be: (1) moballs could explore a much larger expanse of Mars in a much faster fashion, (2) they could explore the difficult terrains such as steep slopes and sand dunes, and (3) they would be self-energy- generating and could work together and move around autonomously. The challenge in designing the structure and the mechanics of the moball would be that it should be sturdy enough to withstand the impact of its initial fall, as well as other impacts from obstacles in its way. A mechanism would be needed that could enable hundreds of moballs to be carried while they would be deflated and compact, then would inflate them just after deploying them to their drop site. Furthermore, the moballs should also be light enough to allow them to move easily over obstacles by force of the wind. They also should have some kind of maneuvering mechanism in place to help them avoid very hazardous sharp objects or events, and to enable them to get closer to the objects of interest. The structure of the moballs was designed so that they would have different layers. The outer layer should comprise a sturdy, yet light, polymer that could withstand both the impact of the initial drop, as well as the impact of the different obstacles it would encounter while traversing the surface of Mars. This polymer should not deteriorate with the 100 K daily temperature swings on Mars. The inner layer should consist of a very light gas such as nitrogen or helium. In terms of maneuvering, six very light weights placed at strategic locations would give moballs the ability to turn, or even hop, over hazardous (e.g., sharp) obstacles, or even initiate a movement (before getting more help from the wind to be carried around) when stuck. Maneuvering would be necessary in order to get closer to objects of interest. If the weights would be allowed to move freely, they could also be used to generate energy.

Davoodi, Faranak↗

DOE OSTI2022

Modernization efforts for the R -Matrix code SAMMY [Abstract]

The R-Matrix code SAMMY is a widely used nuclear data evaluation code focused on the resolved range, which includes corrections for experimental effects. The code is still mostly written in Fortran 77, and uses a memory management system suitable for the time of its initial writing (1984). A modernization effort is under way to bring the code in-line with modern software development practices. A continuous-integration testing framework was added, automating the large existing set of test cases. It is run on every commit. The memory management was updated to current standard practices suitable for modern software analysis tools. The code can be obtained from https://code.ornl.gov/RNSD/SAMMY. The resonance parameters and covariance information are now stored in C++ objects shared by SAMMY and AMPX, the processing code that generates nuclear data libraries for SCALE. This allows for easier maintenance and access to the resonance parameters inside and outside of SAMMY. This feature is already used by accessing and changing parameters in memory in the Bayesian Monte Carlo Evaluation Framework for Cross Sections Nuclear Data and Integral Benchmark Experiments project, Further plans include the switch to the ENDF reading and writing routines in AMPX, as these routines are more robust, easier to maintain, and support more features. Of note here is support for the new GNDS format. Previously it wasn’t easy to share the full covariance matrix for evaluations containing more than one isotope due to limitations on the ENDF format; this is now supported in GNDS. The data are currently available in a binary SAMMY format and can be exported to GNDS to make them more widely available and sharable. The next step will be to use the same resonance processing code at 0K in AMPX and SAMMY as one of the available Reich-Moore R-Matrix formalism. The first step toward this goal is to isolate the reconstruction into a module that takes resonance parameters as its input and does not depend on SAMMY global parameters. This goal has been achieved and it should now be possible to more easily change the resonance formalism and add enhancements as the Phenomenological R-Matrix parameterization of direct, doorway, and compound nuclear reactions discussed elsewhere on this conference. This concerted modernization and enhancement effort provides multiple advantages to the nuclear data community. It will allow parameter optimization using enhanced formalisms, including experimental effects, that better match complex experimental data. Then those evaluated parameters can immediately be passed off to AMPX to be reconstructed with the exact same cross section model and be put into a data library for subsequent testing using SCALE and the Valid Benchmark suite or other suitable benchmark suites.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

DOE OSTI2023

UPC++ v1.0 Programmer’s Guide (Rev. 2023.9.0)

UPC++ is a C++ library that supports Partitioned Global Address Space (PGAS) programming. It is designed for writing efficient, scalable parallel programs on distributed-memory parallel computers. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). The UPC++ control model is single program, multiple-data (SPMD), with each separate constituent process having access to local memory as it would in C++. The PGAS memory model additionally provides one-sided RMA communication to a global address space, which is allocated in shared segments that are distributed over the processes. UPC++ also features Remote Procedure Call (RPC) communication, making it easy to move computation to operate on data that resides on remote processes. UPC++ was designed to support exascale high-performance computing, and the library interfaces and implementation are focused on maximizing scalability. In UPC++, all communication operations are syntactically explicit, which encourages programmers to consider the costs associated with communication and data movement. Moreover, all communication operations are asynchronous by default, encouraging programmers to seek opportunities for overlapping communication latencies with other useful work. UPC++ provides expressive and composable abstractions designed for efficiently managing aggressive use of asynchrony in programs. Together, these design principles are intended to enable programmers to write applications using UPC++ that perform well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

UPC++ v1.0 Programmer’s Guide (Revision 2022.3.0)

UPC++ is a C++ library that supports Partitioned Global Address Space (PGAS) programming. It is designed for writing efficient, scalable parallel programs on distributed-memory parallel computers. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). The UPC++ control model is single program, multiple-data (SPMD), with each separate constituent process having access to local memory as it would in C++. The PGAS memory model additionally provides one-sided RMA communication to a global address space, which is allocated in shared segments that are distributed over the processes. UPC++ also features Remote Procedure Call (RPC) communication, making it easy to move computation to operate on data that resides on remote processes. UPC++ was designed to support exascale high-performance computing, and the library interfaces and implementation are focused on maximizing scalability. In UPC++, all communication operations are syntactically explicit, which encourages programmers to consider the costs associated with communication and data movement. Moreover, all communication operations are asynchronous by default, encouraging programmers to seek opportunities for overlapping communication latencies with other useful work. UPC++ provides expressive and composable abstractions designed for efficiently managing aggressive use of asynchrony in programs. Together, these design principles are intended to enable programmers to write applications using UPC++ that perform well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

UPC++ v1.0 Programmer’s Guide, Revision 2023.3.0

UPC++ is a C++ library that supports Partitioned Global Address Space (PGAS) programming. It is designed for writing efficient, scalable parallel programs on distributed-memory parallel computers. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). The UPC++ control model is single program, multiple-data (SPMD), with each separate constituent process having access to local memory as it would in C++. The PGAS memory model additionally provides one-sided RMA communication to a global address space, which is allocated in shared segments that are distributed over the processes. UPC++ also features Remote Procedure Call (RPC) communication, making it easy to move computation to operate on data that resides on remote processes. UPC++ was designed to support exascale high-performance computing, and the library interfaces and implementation are focused on maximizing scalability. In UPC++, all communication operations are syntactically explicit, which encourages programmers to consider the costs associated with communication and data movement. Moreover, all communication operations are asynchronous by default, encouraging programmers to seek opportunities for overlapping communication latencies with other useful work. UPC++ provides expressive and composable abstractions designed for efficiently managing aggressive use of asynchrony in programs. Together, these design principles are intended to enable programmers to write applications using UPC++ that perform well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

UPC++ v1.0 Programmer’s Guide, Revision 2022.9.0

UPC++ is a C++ library that supports Partitioned Global Address Space (PGAS) programming. It is designed for writing efficient, scalable parallel programs on distributed-memory parallel computers. The key communication facilities in UPC++ are one-sided Remote Memory Access (RMA) and Remote Procedure Call (RPC). The UPC++ control model is single program, multiple-data (SPMD), with each separate constituent process having access to local memory as it would in C++. The PGAS memory model additionally provides one-sided RMA communication to a global address space, which is allocated in shared segments that are distributed over the processes. UPC++ also features Remote Procedure Call (RPC) communication, making it easy to move computation to operate on data that resides on remote processes. UPC++ was designed to support exascale high-performance computing, and the library interfaces and implementation are focused on maximizing scalability. In UPC++, all communication operations are syntactically explicit, which encourages programmers to consider the costs associated with communication and data movement. Moreover, all communication operations are asynchronous by default, encouraging programmers to seek opportunities for overlapping communication latencies with other useful work. UPC++ provides expressive and composable abstractions designed for efficiently managing aggressive use of asynchrony in programs. Together, these design principles are intended to enable programmers to write applications using UPC++ that perform well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1995

Solid-state image sensor with focal-plane digital photon-counting pixel array

A photosensitive layer such as a-Si for a UV/visible wavelength band is provided for low light level imaging with at least a separate CMOS amplifier directly connected to each PIN photodetector diode to provide a focal-plane array of NxN pixels, and preferably a separate photon-counting CMOS circuit directly connected to each CMOS amplifier, although one row of counters may be time shared for reading out the photon flux rate of each diode in the array, together with a buffer memory for storing all rows of the NxN image frame before transfer to suitable storage. All CMOS circuitry is preferably fabricated in the same silicon layer as the PIN photodetector diode for a monolithic structure, but when the wavelength band of interest requires photosensitive material different from silicon, the focal-plane array may be fabricated separately on a different semiconductor layer bump-bonded or otherwise bonded for a virtually monolithic structure with one free terminal of each diode directly connected to the input terminal of its CMOS amplifier and digital counter for integration of the photon flux rate at each photodetector of the array.

Fossum, Eric R.↗

DOE OSTI2021

Demystifying asynchronous I/O Interference in HPC applications

With increasing complexity of HPC workflows, data management services need to perform expensive I/O operations asynchronously in the background, aiming to overlap the I/O with the application runtime. However, this may cause interference due to competition for resources: CPU, memory/network bandwidth. The advent of multi-core architectures has exacerbated this problem, as many I/O operations are issued concurrently, thereby competing not only with the application but also among themselves. Furthermore, the interference patterns can dynamically change as a response to variations in application behavior and I/O subsystems (e.g. multiple users sharing a parallel file system). Without a thorough understanding, I/O operations may perform suboptimally, potentially even worse than in the blocking case. To fill this gap, here we investigate the causes and consequences of interference due to asynchronous I/O on HPC systems. Specifically, we focus on multi-core CPUs and memory bandwidth, isolating the interference due to each resource. Then, we perform an in-depth study to explain the interplay and contention in a variety of resource sharing scenarios such as varying priority and number of background I/O threads and different I/O strategies: sendfile, read/write, mmap/write underlining trade-offs. The insights from this study are important both to enable guided optimizations of existing background I/O, as well as to open new opportunities to design advanced asynchronous I/O strategies.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Unified Memory: GPGPU-Sim/UVM Smart Integration

CPU/GPU heterogeneous compute platforms are an ubiquitous element in computing and a programming model specified for this heterogeneous computing model is important for both performance and programmability. A programming model that exposes the shared, unified, address space between the heterogeneous units is a necessary step in this direction as it removes the burden of explicit data movement from the programmer while maintaining performance. GPU vendors, such as AMD and NVIDIA, have released software-managed runtimes that can provide programmers the illusion of unified CPU and GPU memory by automatically migrating data in and out of the GPU memory. However, this runtime support is not included in GPGPU-Sim, a commonly used framework that models the features of a modern graphics processor that are relevant to non-graphics applications. UVM Smart was developed, which extended GPGPU-Sim 3.x to in- corporate the modeling of on-demand pageing and data migration through the runtime. This report discusses the integration of UVM Smart and GPGPU-Sim 4.0 and the modifications to improve simulation performance and accuracy.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1989

Thin film memory matrix using amorphous and high resistive layers

Memory cells in a matrix are provided by a thin film of amorphous semiconductor material overlayed by a thin film of resistive material. An array of parallel conductors on one side perpendicular to an array of parallel conductors on the other side enable the amorphous semiconductor material to be switched in addressed areas to be switched from a high resistance state to a low resistance state with a predetermined level of electrical energy applied through selected conductors, and thereafter to be read out with a lower level of electrical energy. Each cell may be fabricated in the channel of an MIS field-effect transistor with a separate common gate over each section to enable the memory matrix to be selectively blanked in sections during storing or reading out of data. This allows for time sharing of addressing circuitry for storing and reading out data in a synaptic network, which may be under control of a microprocessor.

Thakoor, Anilkumar P.↗

DOE OSTI2022

Block-structured, equal-workload, multi-grid-nesting interface for the Boussinesq wave model FUNWAVE-TVD (Total Variation Diminishing)

Abstract. We describe the development of a block-structured, equal-CPU-load (central processing unit), multi-grid-nesting interface for the Boussinesq wave model FUNWAVE-TVD (Fully Nonlinear Boussinesq Wave Model with Total Variation Diminishing Solver). The new model framework does not interfere with the core solver, and thus the core program, FUNWAVE-TVD, is still a standalone model used for a single grid. The nesting interface manages the time sequencing and two-way nesting processes between the parent grid and child grid with grid refinement in a hierarchical manner. Workload balance in the MPI-based (message passing interface) parallelization is handled by an equal-load scheme. A strategy of shared array allocation is applied for data management that allows for a large number of nested grids without creating additional memory allocations. Four model tests are conducted to verify the nesting algorithm with assessments of model accuracy and the robustness in the application in modeling transoceanic tsunamis and coastal effects.

Choi, Young-Kwang↗

DOE OSTI2024

Comparison of ML-Based Proxy Modeling Strategies: Lessons Learned from the SMART Initiative

Teams of researchers on Task 5 of the SMART project have developed a variety of modeling architectures to predict subsurface behavior during carbon injection and post-injection periods. One important part of this task was to compare the candidate approaches in terms of accuracy, reliability, speed, and memory use, all using a common set of metrics and visualizations for an “apples to apples” comparison. Dr. Jared Schuetter will share the details of this task, the results that were obtained, and the lessons learned.

Schuetter, Jared↗