Search NASA⌕ Search

SEARCH · Search NASA

Results for “memory efficiency”

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 1,693 records · Page 94

NWChem: scalable parallel computational chemistry

NWChem is a general purpose computational chemistry code specifically designed to run on distributed memory parallel computers. The core functionality of the code focuses on molecular dynamics, Hartree-Fock and density functional theory methods for both plane-wave basis sets as well as Gaussian basis sets, tensor contraction engine based coupled cluster capabilities and combined quantum mechanics/molecular mechanics descriptions. It was realized from the beginning that scalable implementations of these methods required a programming paradigm inherently different from what message passing approaches could offer. In response a global address space library, the Global Array Toolkit, was developed. The programming model it offers is based on using predominantly one-sided communication. This model underpins most of the functionality in NWChem and the power of it is exemplified by the fact that the code scales to tens of thousands of processors. In this paper the core capabilities of NWChem are described as well as their implementation to achieve an efficient computational chemistry code with high parallel scalability. NWChem is a modern, open source, computational chemistry code1 specifically designed for large scale parallel applications2. To meet the challenges of developing efficient, scalable and portable programs of this nature a particular code design was adopted. This code design involved two main features. First of all, the code is build up in a modular fashion so that a large variety of functionality can be integrated easily. Secondly, to facilitate writing complex parallel algorithms the Global Array toolkit was developed. This toolkit allows one to write parallel applications in a shared memory like approach, but offers additional mechanisms to exploit data locality to lower communication overheads. This framework has proven to be very successful in computational chemistry but is applicable to any engineering domain. Within the context created by the features above NWChem has grown into a general purpose computational chemistry code that supports a wide variety of energy expressions and capabilities to calculate properties based there upon. The main energy expressions are classical mechanics force fields, Hartree-Fock and DFT both for finite systems and condensed phase systems, coupled cluster, as well as QM/MM. For most energy expressions single point calculations, geometry optimizations, excited states, and other properties are available. Below we briefly discuss each of the main energy expressions and the critical points involved in scalable implementations thereof.

99 GENERAL AND MISCELLANEOUS↗

Generalized quantum master equations in and out of equilibrium: When can one win?

Generalized quantum master equations (GQMEs) are an important tool in modeling chemical and physical processes. For a large number of problems, it has been shown that exact and approximate quantum dynamics methods can be made dramatically more efficient, and in the latter case more accurate, by proceeding via the GQME formalism. However, there are many situations where utilizing the GQME approach with an approximate method has been observed to return the same dynamics as using that method directly. Here, for systems both in and out of equilibrium, we provide a more detailed understanding of the conditions under which using an approximate method can yield benefits when combined with the GQME formalism. In particular, we demonstrate the necessary manipulations, which are satisfied by exact quantum dynamics, that are required to recast the memory kernel in a form that can be analytically shown to yield the same result as a direct application of the dynamics regardless of the approximation used. By considering the connections between these forms of the kernel, we derive the conditions that approximate methods must satisfy if they are to offer different results when used in conjunction with the GQME formalism. These analytical results thus provide new insights as to when proceeding via the GQME approach can be used to improve the accuracy of simulations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Space and power efficient hybrid counters array

A hybrid counter array device for counting events. The hybrid counter array includes a first counter portion comprising N counter devices, each counter device for receiving signals representing occurrences of events from an event source and providing a first count value corresponding to a lower order bits of the hybrid counter array. The hybrid counter array includes a second counter portion comprising a memory array device having N addressable memory locations in correspondence with the N counter devices, each addressable memory location for storing a second count value representing higher order bits of the hybrid counter array. A control device monitors each of the N counter devices of the first counter portion and initiates updating a value of a corresponding second count value stored at the corresponding addressable memory location in the second counter portion. Thus, a combination of the first and second count values provide an instantaneous measure of number of events received.

Gara, Alan G↗

Space and power efficient hybrid counters array

A hybrid counter array device for counting events. The hybrid counter array includes a first counter portion comprising N counter devices, each counter device for receiving signals representing occurrences of events from an event source and providing a first count value corresponding to a lower order bits of the hybrid counter array. The hybrid counter array includes a second counter portion comprising a memory array device having N addressable memory locations in correspondence with the N counter devices, each addressable memory location for storing a second count value representing higher order bits of the hybrid counter array. A control device monitors each of the N counter devices of the first counter portion and initiates updating a value of a corresponding second count value stored at the corresponding addressable memory location in the second counter portion. Thus, a combination of the first and second count values provide an instantaneous measure of number of events received.

47 OTHER INSTRUMENTATION↗

A Novel Application of Parallel Betweenness Centrality to Power Grid Contingency Analysis

In Energy Management Systems, contingency analysis is commonly performed for identifying and mitigating potentially harmful power grid component failures. The exponentially increasing combinatorial number of failure modes imposes a significant computational burden for massive contingency analysis. It is critical to select a limited set of high-impact contingency cases within the constraint of computing power and time requirements to make it possible for real-time power system vulnerability assessment. In this paper, we present a novel application of parallel betweenness centrality to power grid contingency selection. We cross-validate the proposed method using the model and data of the western US power grid, and implement it on a Cray XMT system - a massively multithreaded architecture - leveraging its advantages for parallel execution of irregular algorithms, such as graph analysis. We achieve a speedup of 55 times (on 64 processors) compared against the single-processor version of the same code running on the Cray XMT. We also compare an OpenMP-based version of the same code running on an HP Superdome shared-memory machine. The performance of the Cray XMT code shows better scalability and resource utilization, and shorter execution time for large-scale power grids. This proposed approach has been evaluated in PNNL’s Electricity Infrastructure Operations Center (EIOC). It is expected to provide a quick and efficient solution to massive contingency selection problems to help power grid operators to identify and mitigate potential widespread cascading power grid failures in real time.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Architecture-Aware Models of AI Engines for High-Performance Matrix Matrix Multiplication

The AI Engine (AIE) architecture, available in systems from mobile SoCs to server-class FPGAs, aims to efficiently execute AI/ML tasks through a two-dimensional array of compute tiles. Previous work on AIEs has explored different approaches to mapping computation across spatial arrays, but the compute kernel running on each tile has not been the focus. Additionally, the AIE-ML architecture introduces memory tiles and omits programmable logic, requiring new approaches to staging and moving data throughout the array. In this work we update analytical models developed for CPUs to produce the design of high performance kernels while introducing new model considerations such as memory structure, throughput, and latency as required by the AIE hardware. We evaluate our models by developing AIE-ML kernels for matrix multiplication in low-precision data types showing performance up to 95% of compute peak for the kernel when data resides in local memory and above 90% of compute peak when data resides in main memory.

Binder, Elliott D. [Carnegie Mellon University, Pi↗

Bloom filters for molecules

Abstract Ultra-large chemical libraries are reaching 10s to 100s of billions of molecules. A challenge for these libraries is to efficiently check if a proposed molecule is present. Here we propose and study Bloom filters for testing if a molecule is present in a set using either string or fingerprint representations. Bloom filters are small enough to hold billions of molecules in just a few GB of memory and check membership in sub milliseconds. We found string representations can have a false positive rate below 1% and require significantly less storage than using fingerprints. Canonical SMILES with Bloom filters with the simple FNV (Fowler-Noll-Voll) hashing function provide fast and accurate membership tests with small memory requirements. We provide a general implementation and specific filters for detecting if a molecule is purchasable, patented, or a natural product according to existing databases at https://github.com/whitead/molbloom .

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Caloric effects in ferroic materials

The fundamentals and applications of ferroic materials$-$ferromagnetic, ferroelectric, and ferroelastic$-$are common subjects discussed in just about every graduate course related to functional materials. Looking beyond today’s traditional uses, such as in permanent magnets, capacitors, and shape-memory alloys, there are worthwhile and interesting questions common to the caloric properties of these ferroic materials. Can ferroic materials be used in a cooling cycle? Why are these materials susceptible to external fields? Which combination of properties is required to make some of them suitable for efficient cooling and heat pumping? We address these questions in this introduction to ferroic cooling, which comprises magnetocaloric, electrocaloric, elastocaloric and barocaloric approaches and combinations thereof (i.e., multicalorics). These are addressed in greater detail in the articles in this issue.

36 MATERIALS SCIENCE↗

Mosaics , The Best of Both Worlds: Analog devices with Digital Spiking Communication to build a Hybrid Neural Network Accelerator

Neuromorphic architectures have seen a resurgence of interest in the past decade owing to 100x-1000x efficiency gain over conventional Von Neumann architectures. Digital neuromorphic chips like Intel's Loihi have shown efficiency gains compared to GPUs and CPUs and can be scaled to build larger systems. Analog neuromorphic architectures promise even further savings in energy efficiency, area, and latency than their digital counterparts. Neuromorphic analog and digital technologies provide both low-power and configurable acceleration of challenging artificial intelligence (AI) algorithms. We present a hybrid analog-digital neuromorphic architecture that can amplify the advantages of both high-density analog memory and spike-based digital communication while mitigating each of the other approaches' limitations.

97 MATHEMATICS AND COMPUTING↗

Unified Language Frontend for Physic-Informed AI/ML

Artificial intelligence and machine learning (AI/ML) are becoming important tools for scientific modeling and simulation as in several other fields such as image analysis and natural language processing. ML techniques can leverage the computing power available in modern systems and reduce the human effort needed to configure experiments, interpret and visualize results, draw conclusions from huge quantities of raw data, and build surrogates for physics based models. Domain scientists in fields like fluid dynamics, microelectronics and chemistry can automate many of their most difficult and repetitive tasks or improve the design times by use of the faster ML-surrogates. However, modern ML and traditional scientific highperformance computing (HPC) tend to use completely different software ecosystems. While ML frameworks like PyTorch and TensorFlow provide Python APIs, most HPC applications and libraries are written in C++. Direct interoperability between the two languages is possible but is tedious and error-prone. In this work, we show that a compiler-based approach can bridge the gap between ML frameworks and scientific software with less developer effort and better efficiency. We use the MLIR (multi-level intermediate representation) ecosystem to compile a pre-trained convolutional neural network (CNN) in PyTorch to freestanding C++ source code in the Kokkos programming model. Kokkos is a programming model widely used in HPC to write portable, shared-memory parallel code that can natively target a variety of CPU and GPU architectures. Our compiler-generated source code can be directly integrated into any Kokkosbased application with no dependencies on Python or cross-language interfaces.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Techniques for writing and reading data on an optical disk which include formation of holographic optical gratings in plural locations on the optical disk

An optical memory for storing and/or reading data on an optical disk. The optical disk incorporates a material in which holographic gratings can be created, and subsequently detected, at plural locations within the disk by an electro-optical head. Creation and detection of holographic gratings with variable diffraction efficiency is possible with the electro-optical head. Multiple holographic gratings can also be created at each one of the plural locations via a beam of light which has a different wavelength or point of focus. These data elements can be read by the electro-optical head using a beam of light sequentially varied in wavelength or point of focus to correspond to the multiple holographic gratings to be recorded.

Liu, Tsuen-Hsi↗

A Screen Space GPGPU Surface LIC Algorithm for Distributed Memory Data Parallel Sort Last Rendering Infrastructures

The surface line integral convolution(LIC) visualization technique produces dense visualization of vector fields on arbitrary surfaces. We present a screen space surface LIC algorithm for use in distributed memory data parallel sort last rendering infrastructures. The motivations for our work are to support analysis of datasets that are too large to fit in the main memory of a single computer and compatibility with prevalent parallel scientific visualization tools such as ParaView and VisIt. By working in screen space using OpenGL we can leverage the computational power of GPUs when they are available and run without them when they are not. We address efficiency and performance issues that arise from the transformation of data from physical to screen space by selecting an alternate screen space domain decomposition. We analyze the algorithm's scaling behavior with and without GPUs on two high performance computing systems using data from turbulent plasma simulations.

Loring, Burlen↗

Dynamic Learning of Correlation Potentials for a Time-Dependent Kohn-Sham System

We develop methods to learn the correlation potential for a time-dependent Kohn-Sham (TDKS) system in one spatial dimension. We start from a low-dimensional two-electron system for which we can numerically solve the time-dependent Schr¨odinger equation; this yields electron densities suitable for training models of the correlation potential. We frame the learning problem as one of optimizing a least-squares objective subject to the constraint that the dynamics obey the TDKS equation. Applying adjoints, we develop efficient methods to compute gradients and thereby learn models of the correlation potential. Our results show that it is possible to learn values of the correlation potential such that the resulting electron densities match ground truth densities. We also show how to learn correlation potential functionals with memory, demonstrating one such model that yields reasonable results for trajectories outside the training set.

97 MATHEMATICS AND COMPUTING↗

Improving Memory Subsystem Performance Using ViVA: Virtual Vector Architecture

The disparity between microprocessor clock frequencies and memory latency is a primary reason why many demanding applications run well below peak achievable performance. Software controlled scratchpad memories, such as the Cell local store, attempt to ameliorate this discrepancy by enabling precise control over memory movement; however, scratchpad technology confronts the programmer and compiler with an unfamiliar and difficult programming model. In this work, we present the Virtual Vector Architecture (ViVA), which combines the memory semantics of vector computers with a software-controlled scratchpad memory in order to provide a more effective and practical approach to latency hiding. ViVA requires minimal changes to the core design and could thus be easily integrated with conventional processor cores. To validate our approach, we implemented ViVA on the Mambo cycle-accurate full system simulator, which was carefully calibrated to match the performance on our underlying PowerPC Apple G5 architecture. Results show that ViVA is able to deliver significant performance benefits over scalar techniques for a variety of memory access patterns as well as two important memory-bound compact kernels, corner turn and sparse matrix-vector multiplication -- achieving 2x-13x improvement compared the scalar version. Overall, our preliminary ViVA exploration points to a promising approach for improving application performance on leading microprocessors with minimal design and complexity costs, in a power efficient manner.

97 MATHEMATICS AND COMPUTING↗

Roadrunner Supercomputer Breaks the Petaflop Barrier

At 3:30 a.m. on May 26, 2008, Memorial Day, the "Roadrunner" supercomputer exceeded a sustained speed of 1 petaflop/s, or 1 million billion calculations per second. The sustained performance makes Roadrunner more than twice as fast as the current number 1 system on the TOP500 list. The best sustained performance to date is 74.5% efficiency, 1.026 petaflop/s. "Petaflop/s" is computer jargon—peta signifying the number 1 followed by 15 zeros (sometimes called a quadrillion) and flop/s meaning "double-precision floating point operations per second." Los Alamos held the fastest supercomputer title in 1993 with the Thinking Machines CM-5, and inaugurated the supercomputer era, assisting in the development of the Cray-1 in 1976. The Laboratory and IBM go all the way back to the first card-programmable calculators, used at Los Alamos in 1949. Los Alamos also housed serial number 1 of the IBM 704 in 1956. The Roadrunner supercomputer, developed by IBM in partnership with the Laboratory and the National Nuclear Security Administration, uses commercially available hardware, including aspects of commercial game console and graphics technologies. Because of its off-the-shelf components, the computer costs significantly less than a one-of-a-kind machine. It also uses a Linux operating system. The secret to its record-breaking performance is a unique hybrid design. Each compute node in this cluster consists of two AMD OpteronTM dual-core processors plus four PowerXCell 8iTM processors used as computational accelerators. The accelerators used in Roadrunner are a special IBM-developed variant of the Cell processor used in the Sony PlayStation® 3. The node-attached Cell accelerators are what make Roadrunner different than typical clusters. The Lab and IBM have been working on Roadrunner since 2006, but collaboration on Cell dates back to 2002. The first phase of the project included delivery of an initial Opteron-only cluster that operates at a speed of 71 teraflop/s. This initial system has been in full production at Los Alamos for almost a year, and Laboratory researchers are using this machine for classified weapons applications. The full-scale Roadrunner machine operates more than 10 times faster than the current installed system. Phase 2 of the Roadrunner project was completed in October 2007. Two external assessments, one by NNSA headquarters and one by an independent team of high performance computing experts, evaluated the machine's potential use for Laboratory applications, the Laboratory's ability to successfully manage the computing system, IBM's ability to deliver the product, and whether computer programs could be adapted to the new system. Based on the positive outcome of the assessments, the Laboratory and NNSA decided to pursue the final phase of the Roadrunner project. The powerful cluster of nodes will process information enabling the Laboratory to use Roadrunner for advanced physics and predictive simulations of complex scientific processes. Weapons science applications that can be processed by Roadrunner are applicable to all three of the U.S. Department of Energy weapons laboratories. The machine will also be well equipped to tackle the intricacies of modeling processes, ranging from the biomolecular to the cosmological. In addition, Los Alamos intends to purchase additional Roadrunner resources to support open science and technology applications. The full-scale Roadrunner system, named in honor of New Mexico's speedy state bird, broke the performance record at IBM's Poughkeepsie, New York, facility. The full machine will be moved to Los Alamos beginning in July and housed in the Nicholas Metropolis Center for Modeling and Simulation. The first computing applications are expected to begin running on the machine in January 2009. The cost of all phases of the Roadrunner project is approximately $120 million. More than 200 Laboratory employees have been involved in this effort. For more information about Roadrunner at Los Alamos National Laboratory, see the Roadrunner website at http://www.lanl.gov/roadrunner.

CELL BASED PROCESSOR↗

Design-Technology Co-Optimization for NVM-based Neuromorphic Processing Elements

An emerging use-case of machine learning (ML) is to train a model on a high-performance system and deploy the trained model on energy-constrained embedded systems. Neuromorphic hardware platforms, which operate on principles of the biological brain, can significantly lower the energy overhead of a machine learning inference task, making these platforms an attractive solution for embedded ML systems. In this paper, we present a design-technology tradeoff analysis to implement such inference tasks on the processing elements (PEs) of a Non-Volatile Memory (NVM)-based neuromorphic hardware. Through detailed circuit-level simulations at scaled process technology nodes, we show the negative impact of technology scaling on the information-processing latency, which impacts the quality-of-service (QoS) of an embedded ML system. At a finer granularity, the latency inside a PE depends on 1) the delay introduced by parasitic components on its current paths, and 2) the varying delay to sense different resistance states of its NVM cells. Based on these two observations, we make the following three contributions. First, on the technology front, we propose an optimization scheme where the NVM resistance state that takes the longest time to sense is set on current paths having the least delay, and vice versa, reducing the average PE latency, which improves the QoS. Second, on the architecture front, we introduce isolation transistors within each PE to partition it into regions that can be individually power-gated, reducing both latency and energy. Finally, on the system-software front, we propose a mechanism to leverage the proposed technological and architectural enhancements when implementing a machine-learning inference task on neuromorphic PEs of the hardware. Evaluations with a recent neuromorphic hardware architecture show that our proposed design-technology co-optimization approach improves both performance and energy efficiency of machine-learning inference tasks without incurring high cost-per-bit.

42 ENGINEERING↗

A Pervasive Parallel Processing Framework for Data Visualization and Analysis at Extreme Scale

The evolution of the computing world from teraflop to petaflop has been relatively effortless, with several of the existing programming models scaling effectively to the petascale. The migration to exascale, however, poses considerable challenges. All industry trends infer that the exascale machine will be built using processors containing hundreds to thousands of cores per chip. It can be inferred that efficient concurrency on exascale machines requires a massive amount of concurrent threads, each performing many operations on a localized piece of data. Currently, visualization libraries and applications are based off what is known as the visualization pipeline. In the pipeline model, algorithms are encapsulated as filters with inputs and outputs. These filters are connected by setting the output of one component to the input of another. Parallelism in the visualization pipeline is achieved by replicating the pipeline for each processing thread. This works well for today’s distributed memory parallel computers but cannot be sustained when operating on processors with thousands of cores. Our project investigates a new visualization framework designed to exhibit the pervasive parallelism necessary for extreme scale machines. Our framework achieves this by defining algorithms in terms of worklets, which are localized stateless operations. Worklets are atomic operations that execute when invoked unlike filters, which execute when a pipeline request occurs. The worklet design allows execution on a massive amount of lightweight threads with minimal overhead. Only with such fine-grained parallelism can we hope to fill the billions of threads we expect will be necessary for efficient computation on an exascale machine.

97 MATHEMATICS AND COMPUTING↗

Study of the modifications needed for efficient operation of NASTRAN on the Control Data Corporation STAR-100 computer

NASA structural analysis (NASTRAN) computer program is operational on three series of third generation computers. The problem and difficulties involved in adapting NASTRAN to a fourth generation computer, namely, the Control Data STAR-100, are discussed. The salient features which distinguish Control Data STAR-100 from third generation computers are hardware vector processing capability and virtual memory. A feasible method is presented for transferring NASTRAN to Control Data STAR-100 system while retaining much of the machine-independent code. Basic matrix operations are noted for optimization for vector processing.

Source record↗