Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “compute in 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 1,549 records · Page 86

Evaluating the Opportunities for Multi-Level Memory – An ASC 2016 L2 Milestone

The next two Advanced Technology platforms for the ASC program will feature complex memory hierarchies – in the Trinity supercomputer being deployed in 2016, Intel’s Knights Landing processors will feature 16GB of on-package, high-bandwidth memory, combined with a larger capacity DDR4 memory and in 2018, the Sierra machine deployed at Lawrence Livermore National Laboratory will feature powerful compute nodes containing POWER9 processors with large capacity memories and an array of coherent GPU accelerators also with high bandwidth memories.

97 MATHEMATICS AND COMPUTING↗

Towards Accurate and Efficient Predictions of Martensitic Transition Temperatures for Shape Memory Alloys from First Principles

Recent rapid progresses in physics theory and computational power have made it possible to predict the martensitic transition temperatures (MTTs) in shape memory alloys (SMAs) from first principles [1-3]. In particular, rigorous while time-consuming thermodynamic integration has been employed to compute the anharmonic phonon free energies, which play a crucial role in determining martensitic phase transitions in SMAs. However, this approach has only been applied to simple binaries, and its accuracy is unsatisfying for certain SMAs such as the most commonly used NiTi. In this work, we report on several new developments to our method that bring first-principles theory and experiment much closer into agreement including the MTT of NiTi, and that improve the computational efficiency significantly. We have applied our refined approach to investigate the Ni0.5Ti0.5-xHfx and PdxNi0.5-xTi0.5 ternaries, and the predicted MTT for each composition is within 100K compared with experiment. We will address various techniques to overcome the difficulty encountered in studying ternaries. Our theoretical approach is expected to be a broadly applicable and predictive theory for designing complex SMAs with desirable properties. [1] J.B. Haskins, A.E. Thompson,and J.W. Lawson, Phys. Rev B 94, 214110 (2016). [2] J.B. HaskinsandJ.W. Lawson, J. App. Phys. 121, 205103 (2017). [3] J.B. Haskins, H. Malmir, S. J. Honrao, L. A. Sandoval, and J.W. Lawson, Acta Materialia 212, 116872 (2017).

Zhigang Wu↗

Towards Accurate and Efficient Predictions of Martensitic Transition Temperatures for Shape Memory Alloys from First Principles

Recent rapid progresses in physics theory and computational power have made it possible to predict the martensitic transition temperatures (MTTs) in shape memory alloys (SMAs) from first principles [1-3]. In particular, rigorous while time-consuming thermodynamic integration has been employed to compute the anharmonic phonon free energies, which play a crucial role in determining martensitic phase transitions in SMAs. However, this approach has only been applied to simple binaries, and its accuracy is unsatisfying for certain SMAs such as the most commonly used NiTi. In this work, we report on several new developments to our method that bring first-principles theory and experiment much closer into agreement including the MTT of NiTi, and that improve the computational efficiency significantly. We have applied our refined approach to investigate the Ni0.5Ti0.5-xHfx and PdxNi0.5-xTi0.5 ternaries, and the predicted MTT for each composition is within 100K compared with experiment. We will address various techniques to overcome the difficulty encountered in studying ternaries. Our theoretical approach is expected to be a broadly applicable and predictive theory for designing complex SMAs with desirable properties.

Zhigang Wu↗

GPU Data Access on Complex Geometries for D3Q19 Lattice Boltzmann Method

GPU performance of the lattice Boltzmann method (LBM) depends heavily on memory access patterns. When LBM is advanced with GPUs on complex computational domains, geometric data is typically accessed indirectly, and lattice data is typically accessed lexicographically in the Structure of Array (SoA) layout. Although there are a variety of existing access patterns beyond the typical choices, no study has yet examined the relative efficacy between them. Here, we compare a suite of memory access schemes via empirical testing and performance modeling. We find strong evidence that semi-direct addressing is the superior addressing scheme for the majority of cases examined: Semi-direct addressing increases computational speed and often reduces memory consumption. For lattice layout, we find that the Collected Structure of Arrays (CSoA) layout outperforms the SoA layout. When compared to state-of-the-art practices, our recommended addressing modifications lead to performance gains between 10-40% across different complex geometries, fluid volume fractions, and resolutions. The modifications also lead to a decrease in memory consumption by as much as 17%. Having discovered these improvements, we examine a highly resolved arterial geometry on a leadership class system. On this system we present the first near-optimal strong results for LBM with arterial geometries run on GPUs. We also demonstrate that the above recommendations remain valid for large scale, many device simulations, which leads to an increased computational speed and average memory usage reductions. To understand these observations, we employ performance modeling which reveals that semi-direct methods outperform indirect methods due to a reduced number of total loads/stores in memory, and that CSoA outperforms SoA and bundling due to improved caching behavior.

Herschlag, Gregory↗

A Brief Summary on Formalizing Parallel Tensor Distributions, Redistributions, and Algorithm Derivations

Large-scale datasets in computational chemistry typically require distributed-memory parallel methods to perform a special operation known as tensor contraction. Tensors are multidimensional arrays, and a tensor contraction is akin to matrix multiplication with special types of permutations. Creating an efficient algorithm and optimized implementation in this domain is complex, tedious, and error-prone. To address this, we develop a notation to express data distributions so that we can apply use automated methods to find optimized implementations for tensor contractions. We consider the spin-adapted coupled cluster singles and doubles method from computational chemistry and use our methodology to produce an efficient implementation. Experiments per- formed on the IBM Blue Gene/Q and Cray XC30 demonstrate impact both improved performance and reduced memory consumption.

97 MATHEMATICS AND COMPUTING↗

Using recordings of brain activity to predict and improve human performance

This project investigated methods for using recordings of event-related potentials (ERPs) to predict and improve human memory performance. In a series of experiments, the ERPs related to good and poor memory performance were characterized and a computational model was developed in which the ERPs recorded while a person studied words were used to predict whether that person would be a high performer or a low performer on a subsequent memory test. These experiments provide the first demonstration of ERPs used to predict subsequent memory performance. Additional experiments investigated the effects of different types of memory training (learning a memory strategy such as mental imagery or increasing memory capacity through working memory training) on memory performance and the associated brain activity.

Matzen, Laura E.↗

A Byzantine resilient processor with an encoded fault-tolerant shared memory

The memory requirements for ultra-reliable computers are expected to increase due to future increases in mission functionality and operating-system requirements. This increase will have a negative effect on the reliability and cost of the system. Increased memory size will also reduce the ability to reintegrate a channel after a transient fault, since the time required to reintegrate a channel in a conventional fault-tolerant processor is dominated by memory realignment time. A Byzantine Resilient Fault-Tolerant Processor with Fault-Tolerant Shared Memory (FTP/FTSM) is presented as a solution to these problems. The FTSM uses an encoded memory system, which reduces the memory requirement by one-half compared to a conventional quad-FTP design. This increases the reliability and decreases the cost of the system. The realignment problem is also addressed by the FTSM. Because any single error is corrected upon a read from the FTSM, a faulty channel's corrupted memory does not need realignment before reintegration of the faulty channel. A combination of correct-on-access and background scrubbing is proposed to prevent the accumulation of transient errors in the memory. With a hardware-implemented scrubber, the scrubbing cycle time, and therefore the memory fault latency, can be upper-bounded at a small value. This technique increases the reliability of the memory system and facilitates validation of its reliability model.

Butler, Bryan↗

Scalable Memory Registration for High-Performance Networks Using Helper Threads

Remote DMA (RDMA) enables high performance networks to reduce data copying between an application and the operating system (OS). However RDMA operations in some high performance networks require communication memory explicitly registered with the network adapter and pinned by the OS. Memory registration and pinning limits the flexibility of the memory system and reduces the amount of memory that user processes can allocate. These issues become more significant on multicore platforms, since registered memory demand grows linearly with the number of processor cores. In this paper we propose a new memory registration/deregistration strategy to reduce registered memory on multicore architectures for HPC applications. We hide the cost of dynamic memory management by offloading all dynamic memory registration and deregistration requests to a dedicated memory management helper thread. We investigate design policies and performance implications of the helper thread approach. We evaluate our framework with the NAS parallel benchmarks, for which our registration scheme significantly reduces the registered memory (23.62% on average and up to 49.39%) and avoids memory registration/deregistration costs for reused communication memory. We show that our system enables the execution of problem sizes that could not complete under existing memory registration strategies.

99 GENERAL AND MISCELLANEOUS↗

Implementing Molecular Dynamics on Hybrid High Performance Computers - Three-Body Potentials

The use of coprocessors or accelerators such as graphics processing units (GPUs) has become popular in scientific computing applications due to their low cost, impressive floating-point capabilities, high memory bandwidth, and low electrical power re- quirements. Hybrid high-performance computers, defined as machines with nodes containing more than one type of floating-point processor (e.g. CPU and GPU), are now becoming more prevalent due to these advantages. Although there has been extensive research into methods to efficiently use accelerators to improve the performance of molecular dynamics (MD) employing pairwise potential energy models, little is reported in the literature for models that include many-body effects. 3-body terms are required for many popular potentials such as MEAM, Tersoff, REBO, AIREBO, Stillinger-Weber, Bond-Order Potentials, and others. Because the per-atom simulation times are much higher for models incorporating 3-body terms, there is a clear need for efficient algo- rithms usable on hybrid high performance computers. Here, we report a shared-memory force-decomposition for 3-body potentials that avoids memory conflicts to allow for a deterministic code with substantial performance improvements on hybrid machines. We describe modifications necessary for use in distributed memory MD codes and show results for the simulation of water with Stillinger-Weber on the hybrid Titan supercomputer. We compare performance of the 3-body model to the SPC/E water model when using accelerators. Finally, we demonstrate that our approach can attain a speedup of 5.1 with acceleration on Titan for production simulations to study water droplet freezing on a surface.

Brown, W Michael↗

Accelerating geostatistical modeling using geostatistics-informed machine Learning

Ordinary Kriging (OK) is a popular geostatistical algorithm for spatial interpolation and estimation. The computational complexity of OK changes quadratically and cubically for memory and speed, respectively, given the number of data. Therefore, it is computationally intensive and also challenging to process a large set of data, especially in three-dimensional (3D) cases. This paper develops a geostatistics-informed machine learning (GIML) model to improve the efficiency of OK by reducing the number of points required to be estimated using OK. Specifically, only a very few of the unknown points are estimated by OK to get the weights and estimations, which are used as the training dataset. Moreover, the governing equations of OK are used to guide our proposed machine learning to better reproduce the spatial distributions. Our results show that the proposed GIML can reduce the computational time of OK by at least one order of magnitude. The effectiveness of the GIML is evaluated and compared using a 2D case. Furthermore, we demonstrate its efficiency and robustness by considering a different number of training samples on various 3D simulation grids.

58 GEOSCIENCES↗

Oineus v1.0

A library for multi-threaded computation of persistence diagrams. The algorithm for computing persistence diagrams in a lock-free manner was published in the 'Towards Lock-free Persistent Homology' (D. Morozov, A. Nigmetov, Brief Announcement: SPAA 2020); it scales better than the only other shared-memory parallel implementation of persistent homology computation PHAT. Library includes python bindings to compute persistence diagrams and their vectorizations for lower-star filtrations on grid data. It is intended to be used by scientists working on Topological Data Analysis and its applications in different areas.

Nigmetov, Arnur↗

Boundary element analysis on vector and parallel computers

Boundary element analysis (BEA) can be characterized as a numerical technique that generally shifts the computational burden in the analysis toward numerical integration and the solution of nonsymmetric and either dense or blocked sparse systems of algebraic equations. Researchers have explored the concept that the fundamental characteristics of BEA can be exploited to generate effective implementations on vector and parallel computers. In this paper, the results of some of these investigations are discussed. The performance of overall algorithms for BEA on vector supercomputers, massively data parallel single instruction multiple data (SIMD), and relatively fine grained distributed memory multiple instruction multiple data (MIMD) computer systems is described. Some general trends and conclusions are discussed, along with indications of future developments that may prove fruitful in this regard.

Kane, J. H.↗

Advances in Parallel Electromagnetic Codes for Accelerator Science and Development

Over a decade of concerted effort in code development for accelerator applications has resulted in a new set of electromagnetic codes which are based on higher-order finite elements for superior geometry fidelity and better solution accuracy. SLAC's ACE3P code suite is designed to harness the power of massively parallel computers to tackle large complex problems with the increased memory and solve them at greater speed. The US DOE supports the computational science R&D under the SciDAC project to improve the scalability of ACE3P, and provides the high performance computing resources needed for the applications. This paper summarizes the advances in the ACE3P set of codes, explains the capabilities of the modules, and presents results from selected applications covering a range of problems in accelerator science and development important to the Office of Science.

43 PARTICLE ACCELERATORS↗

Multi-core and Many-core Shared-memory Parallel Raycasting Volume Rendering Optimization and Tuning

Given the computing industry trend of increasing processing capacity by adding more cores to a chip, the focus of this work is tuning the performance of a staple visualization algorithm, raycasting volume rendering, for shared-memory parallelism on multi-core CPUs and many-core GPUs. Our approach is to vary tunable algorithmic settings, along with known algorithmic optimizations and two different memory layouts, and measure performance in terms of absolute runtime and L2 memory cache misses. Our results indicate there is a wide variation in runtime performance on all platforms, as much as 254% for the tunable parameters we test on multi-core CPUs and 265% on many-core GPUs, and the optimal configurations vary across platforms, often in a non-obvious way. For example, our results indicate the optimal configurations on the GPU occur at a crossover point between those that maintain good cache utilization and those that saturate computational throughput. This result is likely to be extremely difficult to predict with an empirical performance model for this particular algorithm because it has an unstructured memory access pattern that varies locally for individual rays and globally for the selected viewpoint. Our results also show that optimal parameters on modern architectures are markedly different from those in previous studies run on older architectures. And, given the dramatic performance variation across platforms for both optimal algorithm settings and performance results, there is a clear benefit for production visualization and analysis codes to adopt a strategy for performance optimization through auto-tuning. These benefits will likely become more pronounced in the future as the number of cores per chip and the cost of moving data through the memory hierarchy both increase.

97 MATHEMATICS AND COMPUTING↗

Module comprising IC memory stack dedicated to and structurally combined with an IC microprocessor chip

A computer module is disclosed in which a stack of glued together IC memory chips is structurally integrated with a microprocessor chip. The memory provided by the stack is dedicated to the microprocessor chip. The microprocessor and its memory stack may be connected either by glue and/or by solder bumps. The solder bumps can perform three functions--electrical interconnection, mechanical connection, and heat transfer. The electrical connections in some versions are provided by wire bonding.

Carson, John C.↗

Logic in Memory Emulator

Logic in Memory Emulator (LiME) is a hardware/software tool specially designed for memory system evaluation and experiment. Emerging memories display a wide range of bandwidths, latencies, and capacities, making it challenging for the computer architect to navigate the design space of potential memory configurations, and for the application developer to assess performance implications of using such memories. With the LiME framework, architectural ideas can be prototyped in great detail yet with sufficient performance to support realistic evaluation on long running applications. LiME consists of two fundamental components: 1) the hardware and OS infrastructure for the emulator, and 2) a suite of benchmark applications to assist in characterizing the performance of current and future computer architectures. Some of the applications have been collected from other open source projects. Uses: Logging, replay and analysis of an application's memory behavior Evaluate impact of emerging memory technology on application performance. Emulate complex memory interactions in whole applications orders of magnitude faster than software simulation. Emulate acceleration hardware co-located with the memory subsystem. Features: Capture and log external memory accesses to a separate off-chip memory device without affecting application execution. Memory traces include the address, length, timestamp, and optionally the data for each transaction. Captured trace data can be saved to an SD card for off-line analysis. Configure a wide range of memory latencies in sub-nanosecond increments that encompass highbandwidth and storage class memories. Specify regions of interest (ROI) in applications to reduce the amount of trace data captured for analysis. Currently supports execution on Xilinx Zynq SoC which integrates an ARM processor with FPGA logic on a single device. Applications can be run under Linux or in bare metal mode on the ARM cores.

Jain, AbhishekK↗

Job-mix modeling and system analysis of an aerospace multiprocessor.

An aerospace guidance computer organization, consisting of multiple processors and memory units attached to a central time-multiplexed data bus, is described. A job mix for this type of computer is obtained by analysis of Apollo mission programs. Multiprocessor performance is then analyzed using: 1) queuing theory, under certain 'limiting case' assumptions; 2) Markov process methods; and 3) system simulation. Results of the analyses indicate: 1) Markov process analysis is a useful and efficient predictor of simulation results; 2) efficient job execution is not seriously impaired even when the system is so overloaded that new jobs are inordinately delayed in starting; 3) job scheduling is significant in determining system performance; and 4) a system having many slow processors may or may not perform better than a system of equal power having few fast processors, but will not perform significantly worse.

Mallach, E. G.↗

Parallel computational fluid dynamics - Implementations and results

The present volume on parallel CFD discusses implementations on parallel machines, numerical algorithms for parallel CFD, and performance evaluation and computer science issues. Attention is given to a parallel algorithm for compressible flows through rotor-stator combinations, a massively parallel Euler solver for unstructured grids, a fast scheme to analyze 3D disk airflow on a parallel computer, and a block implicit multigrid solution of the Euler equations. Topics addressed include a 3D ADI algorithm on distributed memory multiprocessors, clustered element-by-element computations for fluid flow, hypercube FFT and the Fourier pseudospectral method, and an investigation of parallel iterative algorithms for CFD. Also discussed are fluid dynamics using interface methods on parallel processors, sorting for particle flow simulation on the connection machine, a large grain mapping method, and efforts toward a Teraflops capability for CFD.

Simon, Horst D.↗