Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “COMPUTER MEMORY”

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 73 records · Page 4

Rapid solution of large-scale systems of equations

The analysis and design of complex aerospace structures requires the rapid solution of large systems of linear and nonlinear equations, eigenvalue extraction for buckling, vibration and flutter modes, structural optimization and design sensitivity calculation. Computers with multiple processors and vector capabilities can offer substantial computational advantages over traditional scalar computer for these analyses. These computers fall into two categories: shared memory computers and distributed memory computers. This presentation covers general-purpose, highly efficient algorithms for generation/assembly or element matrices, solution of systems of linear and nonlinear equations, eigenvalue and design sensitivity analysis and optimization. All algorithms are coded in FORTRAN for shared memory computers and many are adapted to distributed memory computers. The capability and numerical performance of these algorithms will be addressed.

Storaasli, Olaf O.↗

Compact Ferroelectric Programmable Majority Gate for Compute-in-Memory Applications

In this study, a compact and novel ferroelectric (FE) programmable majority gate is proposed and its novel application in Binary Neural Network (BNNs) is investigated. We demonstrate: i) by integrating N metal-ferroelectric-metal (MFM) capacitors on the gate of a transistor (1T-N-MFM structure), a nonvolatile and programmable majority (MAJ) gate that performs MAJ of AND between the gate input and polarization is realized; ii) validation the functionality of our 3-input MAJ of AND gate through comprehensive theoretical and experimental investigations; iii) a compact implementation of 3-input MAJ of XNOR gate that leverages only five of our 3-input MAJ of AND gates connected in parallel; iv) application of MAJ of XNOR gates to replace the XNOR gates and the first layer of the adder tree in the BNNs for up to 21x area saving on top of eliminating the energy-hungry memory accesses due to the compute-in-memory nature.

97 MATHEMATICS AND COMPUTING↗

Shifting Between Compute and Memory Bounds: A Compression-Enabled Roofline Model

In the evolving landscape of high-performance computing, especially to fight the end of Moore’s Law and Dennard’s Scaling, the ability to shift between compute-bound and memory-bound states is critical for enhancing adaptability and flexibility to diverse system and domain-specific architectures. Such capability is vital for optimizing performance across distinguished hardware configurations, such as accelerators, memory hierarchies, and cache systems. Despite that ad hoc optimization techniques, such as compressed/approximate computation, have been enabled for compute-/data-intensive computing for improved performance in distinct hardware settings, there lacks an understanding of 1) the rational behind performance improvement; 2) capability of different optimizations; 3) what optimization to respond to specific computational and memory demands. This work proposes a compression-enabled roofline model to facilitate this adaptability with data compression techniques to balance and transform between computational and memory demands. This model enables applications to adjust in response to the specific strengths and limitations of the underlying hardware and system to optimize resource utilization. The effectiveness of this approach is demonstrated with matrix multiplication kernels on different input sizes, with turning on/off various compression techniques, including 1) low-precision floating point; 2) sparse matrix formulation; and 3) compressed arrays with ZFP. By reducing memory transfer volumes and cache misses and increasing data locality and computational intensity through compression, the specific roofline model can transform between compute and memory bounds to align more efficiently with system capabilities. This advancement not only improves overall performance but also maximizes adaptability in diverse computing environments.

Naraparaju, Ramasoumya [University of Washington]↗

Synchronization for CXL Based Memory

Compute Express Link (CXL) is an important emerging standard for disaggregated memory. While this standard provisions coherency across numerous hosts and devices, implementing hardware support for type three devices is challenging. In this work, we look at the overhead of software synchronization and using software-based coherency. Moreover, we discuss the limits of software-based coherency in fully expressing modern synchronization techniques for a CXL-based disaggregate memory system. We demonstrate our approach using a CXL hardware prototype and running a version of the famous Peterson Lock (enhanced to run with more than two threads). We analyze its performance and share how more advanced synchronization techniques might interact with software-based coherence CXL hardware and program execution models.

High Performance Computing (HPC)↗

Injecting Artificial Memory Errors Into a Running Computer Program

Single-event upsets (SEUs) or bitflips are computer memory errors caused by radiation. BITFLIPS (Basic Instrumentation Tool for Fault Localized Injection of Probabilistic SEUs) is a computer program that deliberately injects SEUs into another computer program, while the latter is running, for the purpose of evaluating the fault tolerance of that program. BITFLIPS was written as a plug-in extension of the open-source Valgrind debugging and profiling software. BITFLIPS can inject SEUs into any program that can be run on the Linux operating system, without needing to modify the program s source code. Further, if access to the original program source code is available, BITFLIPS offers fine-grained control over exactly when and which areas of memory (as specified via program variables) will be subjected to SEUs. The rate of injection of SEUs is controlled by specifying either a fault probability or a fault rate based on memory size and radiation exposure time, in units of SEUs per byte per second. BITFLIPS can also log each SEU that it injects and, if program source code is available, report the magnitude of effect of the SEU on a floating-point value or other program variable.

Bornstein, Benjamin J.↗

Scalable Cloning on Large-Scale GPU Platforms with Application to Time-Stepped Simulations on Grids

Cloning is a technique to efficiently simulate a tree of multiple what-if scenarios that are unraveled during the course of a base simulation. However, cloned execution is highly challenging to realize on large, distributed memory computing platforms, due to the dynamic nature of the computational load across clones, and due to the complex dependencies spanning the clone tree. In this paper, we present the conceptual simulation framework, algorithmic foundations, and runtime interface of CloneX, a new system we designed for scalable simulation cloning. It efficiently and dynamically creates whole logical copies of a dynamic tree of simulations across a large parallel system without full physical duplication of computation and memory. The performance of a prototype implementation executed on up to 1,024 graphical processing units of a supercomputing system has been evaluated with three benchmarks—heat diffusion, forest fire, and disease propagation models—delivering a speed up of over two orders of magnitude compared to replicated runs. Finally, the results demonstrate a significantly faster and scalable way to execute many what-if scenario ensembles of large simulations via cloning using the CloneX interface.

97 MATHEMATICS AND COMPUTING↗

The effects of dose calculation resolution on dose accuracy for radiation therapy treatments of the lung. Part II. A comparison of dose distributions from an explicit lung model to dose distributions derived from a CT representation

Purpose: Due to limitations in computer memory and computation time, typical radiation therapy treatments are calculated with a voxel dimension on the order of several millimeters. The anatomy below this practical resolution is approximated as a homogeneous region uniform in atomic composition and density. The purpose of this article is to examine whether the exclusion of anatomic structure below the practical dose calculation resolution produces deviations in the resulting dose distributions. Methods: EGSnrc calculated dose distributions from the BRANCH lung model of Part I are compared and contrasted to dose distributions from a CT representation of the same BRANCH model for three different phases of the respiration cycle. Results: The exclusion of branching structures below a CT resolution of 1x1x2 mm{sup 3} resulted in a deviation in dose. The deviation in dose was as high as 14% but was localized around the branching structures. There was no significant variation in the dose deviation as a function of either field size or lung density. Conclusions: The exclusion of explicit branching structures of the lung in a CT representation creates localized deviations in dose. To ensure accurate dose calculations, CT resolution must be increased.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Climatespark: an In-Memory Distributed Computing Framework for Big Climate Data Analytics

The unprecedented growth of climate data creates new opportunities for climate studies, and yet big climate data pose a grand challenge to climatologists to efficiently manage and analyze big data. The complexity of climate data content and analytical algorithms increases the difficulty of implementing algorithms on high performance computing systems. This paper proposes an in-memory, distributed computing framework, ClimateSpark, to facilitate complex big data analytics and time-consuming computational tasks. Chunking data structure improves parallel I/O efficiency, while a spatiotemporal index is built for the chunks to avoid unnecessary data reading and preprocessing. An integrated, multi-dimensional, array-based data model (ClimateRDD) and ETL operations are developed to address big climate data variety by integrating the processing components of the climate data lifecycle. ClimateSpark utilizes Spark SQL and Apache Zeppelin to develop a web portal to facilitate the interaction among climatologists, climate data, analytic operations and computing resources (e.g., using SQL query and Scala/Python notebook). Experimental results show that ClimateSpark conducts different spatiotemporal data queries/analytics with high efficiency and data locality. ClimateSpark is easily adaptable to other big multiple- dimensional, array-based datasets in various geoscience domains.

Hu, Fei↗

Los Alamos works on a biologically realistic computer network

Brain neuroscientists and computer scientists call this field neuromimetic computing—building computers inspired by how the cerebral cortex of the brain works. Its cortical processes rely on billions of small biological “switches,” called neurons. To learn, they store and process information. Using an approach called neural networks, researchers are developing computers that simulate neurons and their interconnections. Then computers can learn about their surroundings, interpret data, and make predictions based on it. In practice, however, researchers attempting to simulate neural processing at anything close to the scale and complexity of the brain’s cortical circuits have been stymied by limitations on computer memory and computational power. All that has changed with the new Trinity supercomputer at Los Alamos, which became fully operational in mid-2017. Trinity has unique capabilities designed for the stockpile stewardship mission, which includes highly complex nuclear simulations that have replaced the testing of nuclear weapons. All this capability means Trinity allows a fundamentally different approach to large-scale cortical simulations, enabling an unprecedented leap in the ability to model neural processing. To test that capability on a limited-scale problem, computer scientists and neuroscientists at Los Alamos created a “sparse prediction machine” that runs on a neural network on Trinity. A sparse prediction machine is designed to work like the brain: researchers expose it to data—in this case, various videos of a car driving down a road—without labeling the data in any way. Then the program sorts through that data frame by frame, focuses on the important information, and develops a prediction about the car’s motion. With Trinity’s power, the Los Alamos team simulates the way a brain handles information in its neurons but uses the fewest neurons at any given moment to explain the information at hand. That’s the “sparse” part, and it makes the brain very efficient—and, hopefully, a computer more efficient, too.

97 MATHEMATICS AND COMPUTING↗

Computational processor-in-memory with enhanced strided memory access

A computational memory for a computer. The memory includes a memory bank having a selected-row buffer and being configured to store records up to a number, K. The memory also includes an accumulator connected to the memory bank, the accumulator configured to store up to K records. The memory also includes an arithmetic and logic unit (ALU) connected to the accumulator and to the selected row buffer of the memory bank, the ALU having an indirect network of 2K ports for reading and writing records in the memory bank and the accumulator, and the ALU further physically configured to operate as a sorting network. The memory also includes a controller connected to the memory bank, the ALU, and the accumulator, the controller being hardware configured to direct operation of the ALU.

DeBenedictis, Erik↗

Performance Evaluation of Remote Memory Access (RMA) Programming on Shared Memory Parallel Computers

The purpose of this study is to evaluate the feasibility of remote memory access (RMA) programming on shared memory parallel computers. We discuss different RMA based implementations of selected CFD application benchmark kernels and compare them to corresponding message passing based codes. For the message-passing implementation we use MPI point-to-point and global communication routines. For the RMA based approach we consider two different libraries supporting this programming model. One is a shared memory parallelization library (SMPlib) developed at NASA Ames, the other is the MPI-2 extensions to the MPI Standard. We give timing comparisons for the different implementation strategies and discuss the performance.

Jin, Hao-Qiang↗

Compute in‐Memory with Non‐Volatile Elements for Neural Networks: A Review from a Co‐Design Perspective

Abstract Deep learning has become ubiquitous, touching daily lives across the globe. Today, traditional computer architectures are stressed to their limits in efficiently executing the growing complexity of data and models. Compute‐in‐memory (CIM) can potentially play an important role in developing efficient hardware solutions that reduce data movement from compute‐unit to memory, known as the von Neumann bottleneck. At its heart is a cross‐bar architecture with nodal non‐volatile‐memory elements that performs an analog multiply‐and‐accumulate operation, enabling the matrix‐vector‐multiplications repeatedly used in all neural network workloads. The memory materials can significantly influence final system‐level characteristics and chip performance, including speed, power, and classification accuracy. With an over‐arching co‐design viewpoint, this review assesses the use of cross‐bar based CIM for neural networks, connecting the material properties and the associated design constraints and demands to application, architecture, and performance. Both digital and analog memory are considered, assessing the status for training and inference, and providing metrics for the collective set of properties non‐volatile memory materials will need to demonstrate for a successful CIM technology.

36 MATERIALS SCIENCE↗

(U) Computation acceleration using dynamic memory

Many computational applications require the repeated use of quantities, whose calculations can be expensive. In order to speed up the overall execution of the program, it is often advantageous to replace computation with extra memory usage. In this approach, computed values are stored and then, when they are needed again, they are quickly retrieved from memory rather than being calculated again at great cost. Sometimes, however, the precise amount of memory needed to store such a collection is not known in advance, and only emerges in the course of running the calculation. One problem accompanying such a situation is wasted memory space in overdimensioned (and possibly sparse) arrays. Another issue is the overhead of copying existing values to a new, larger memory space, if the original allocation turns out to be insufficient. In order to handle these runtime problems, the programmer therefore has the extra task of addressing them in the code.

97 MATHEMATICS AND COMPUTING↗

The Science of Computing: Virtual Memory

In the March-April issue, I described how a computer's storage system is organized as a hierarchy consisting of cache, main memory, and secondary memory (e.g., disk). The cache and main memory form a subsystem that functions like main memory but attains speeds approaching cache. What happens if a program and its data are too large for the main memory? This is not a frivolous question. Every generation of computer users has been frustrated by insufficient memory. A new line of computers may have sufficient storage for the computations of its predecessor, but new programs will soon exhaust its capacity. In 1960, a longrange planning committee at MIT dared to dream of a computer with 1 million words of main memory. In 1985, the Cray-2 was delivered with 256 million words. Computational physicists dream of computers with 1 billion words. Computer architects have done an outstanding job of enlarging main memories yet they have never kept up with demand. Only the shortsighted believe they can.

Denning, Peter J.↗

Distributed-memory lattice H-matrix factorization

We parallelize the LU factorization of a hierarchical low-rank matrix (H-matrix) on a distributed-memory computer. This is much more difficult than the H-matrix-vector multiplication due to the dataflow of the factorization, and it is much harder than the parallelization of a dense matrix factorization due to the irregular hierarchical block structure of the matrix. Block low-rank (BLR) format gets rid of the hierarchy and simplifies the parallelization, often increasing concurrency. However, this comes at a price of losing the near-linear complexity of the H-matrix factorization. In this work, we propose to factorize the matrix using a “lattice H-matrix” format that generalizes the BLR format by storing each of the blocks (both diagonals and off-diagonals) in the H-matrix format. These blocks stored in the H-matrix format are referred to as lattices. Thus, this lattice format aims to combine the parallel scalability of BLR factorization with the near-linear complexity of H-matrix factorization. We first compare factorization performances using the H-matrix, BLR, and lattice H-matrix formats under various conditions on a shared-memory computer. Our performance results show that the lattice format has storage and computational complexities similar to those of the H-matrix format, and hence a much lower cost of factorization than BLR. In conclusion, we then compare the BLR and lattice H-matrix factorization on distributed-memory computers. Our performance results demonstrate that compared with BLR, the lattice format with the lower cost of factorization may lead to faster factorization on the distributed-memory computer.

97 MATHEMATICS AND COMPUTING↗

Nanoimaging of Organic Charge Retention Effects: Implications for Nonvolatile Memory, Neuromorphic Computing, and High Dielectric Breakdown Devices

While a large variety of organic and molecular materials have been found to exhibit charge memory effects, the underlying mechanism is not well-understood, which hinders rational device design. Here, we study the charge retention mechanism of a nanoscale memory system, an organic monolayer on a silicon substrate, with Au nanoparticles on top serving as the electrical contact. Combining scanning probe imaging/manipulation and density functional simulations, we observe stable charge retention effects in the system and attributed it to polaron effects at the amine functional groups. Finally, our findings can pave the way for applications in nonvolatile memory, neuromorphic computing, and high dielectric breakdown devices.

36 MATERIALS SCIENCE↗

Simulation of radiation effects on three-dimensional computer optical memories

A model was developed to simulate the effects of heavy charged-particle (HCP) radiation on the information stored in three-dimensional computer optical memories. The model is based on (i) the HCP track radial dose distribution, (ii) the spatial and temporal distribution of temperature in the track, (iii) the matrix-specific radiation-induced changes that will affect the response, and (iv) the kinetics of transition of photochromic molecules from the colored to the colorless isomeric form (bit flip). It is shown that information stored in a volume of several nanometers radius around the particle's track axis may be lost. The magnitude of the effect is dependent on the particle's track structure.

Non-NASA Center↗