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 1,585 records · Page 88

Extreme Scale De Novo Metagenome Assembly

Metagenome assembly is the process of transforming a set of short, overlapping, and potentially erroneous DNA segments from environmental samples into the accurate representation of the underlying microbiomes's genomes. State-of-the-art tools require big shared memory machines and cannot handle contemporary metagenome datasets that exceed Terabytes in size. In this paper, we introduce the MetaHipMer pipeline, a high-quality and high-performance metagenome assembler that employs an iterative de Bruijn graph approach. MetaHipMer leverages a specialized scaffolding algorithm that produces long scaffolds and accommodates the idiosyncrasies of metagenomes. MetaHipMer is end-to-end parallelized using the Unified Parallel C language and therefore can run seamlessly on shared and distributed-memory systems. Experimental results show that MetaHipMer matches or outperforms the state-of-the-art tools in terms of accuracy. Ultimately, MetaHipMer scales efficiently to large concurrencies and is able to assemble previously intractable grand challenge metagenomes. We demonstrate the unprecedented capability of MetaHipMer by computing the first full assembly of the Twitchell Wetlands dataset, consisting of 7.5 billion reads - size 2.6 TBytes.

97 MATHEMATICS AND COMPUTING↗

Command and data handling of science signals on Spacelab

The Orbiter Avionics and the Spacelab Command and Data Management System (CDMS) combine to provide a relatively complete command, control, and data handling service to the instrument complement during a Shuttle Sortie Mission. The Spacelab CDMS services the instruments and the Orbiter in turn services the Spacelab. The CDMS computer system includes three computers, two I/O units, a mass memory, and a variable number of remote acquisition units. Attention is given to the CDMS high rate multiplexer, CDMS tape recorders, closed circuit television for the visual monitoring of payload bay and cabin area activities, methods of science data acquisition, questions of transmission and recording, CDMS experiment computer usage, and experiment electronics.

Mccain, H. G.↗

Concurrent file operations in a high performance FORTRAN

Distributed memory multiprocessor systems can provide the computing power necessary for large scale scientific applications. A critical performance issue for a number of these applications is the efficient transfer of data to secondary storage. Recently several research groups have proposed FORTRAN language extensions for exploiting the data parallelism of such scientific codes on distributed memory architectures. However, few of these high performance FORTRAN's provide appropriate constructs for controlling the use of the parallel I/O capabilities of modern multiprocessing machines. In this paper, we propose constructs to specify I/O operations for distributed data structures in the context of Vienna Fortran. These operations can be used by the programmer to provide information which can help the compiler and runtime environment make the most efficient use of the I/O subsystem.

Brezany, Peter↗

Distributed memory parallel Markov random fields using graph partitioning

Markov random fields (MRF) based algorithms have attracted a large amount of interest in image analysis due to their ability to exploit contextual information about data. Image data generated by experimental facilities, though, continues to grow larger and more complex, making it more difficult to analyze in a reasonable amount of time. Applying image processing algorithms to large datasets requires alternative approaches to circumvent performance problems. Aiming to provide scientists with a new tool to recover valuable information from such datasets, we developed a general purpose distributed memory parallel MRF-based image analysis framework (MPI-PMRF). MPI-PMRF overcomes performance and memory limitations by distributing data and computations across processors. The proposed approach was successfully tested with synthetic and experimental datasets. Additionally, the performance of the MPI-PMRF framework is analyzed through a detailed scalability study. We show that a performance increase is obtained while maintaining an accuracy of the segmentation results higher than 98%. The contributions of this paper are: (a) development of a distributed memory MRF framework; (b) measurement of the performance increase of the proposed approach; (c) verification of segmentation accuracy in both synthetic and experimental, real-world datasets

97 MATHEMATICS AND COMPUTING↗

Scalable Triadic Analysis of Large-Scale Graphs: Multi-Core vs. Multi-Processor vs. Multi-Threaded Shared Memory Architectures

Triadic analysis encompasses a useful set of graph mining methods that is centered on the concept of a triad, which is a subgraph of three nodes and the configuration of directed edges across the nodes. Such methods are often applied in the social sciences as well as many other diverse fields. Triadic methods commonly operate on a triad census that counts the number of triads of every possible edge configuration in a graph. Like other graph algorithms, triadic census algorithms do not scale well when graphs reach tens of millions to billions of nodes. To enable the triadic analysis of large-scale graphs, we developed and optimized a triad census algorithm to efficiently execute on shared memory architectures. We will retrace the development and evolution of a parallel triad census algorithm. Over the course of several versions, we continually adapted the code’s data structures and program logic to expose more opportunities to exploit parallelism on shared memory that would translate into improved computational performance. We will recall the critical steps and modifications that occurred during code development and optimization. Furthermore, we will compare the performances of triad census algorithm versions on three specific systems: Cray XMT, HP Superdome, and AMD multi-core NUMA machine. These three systems have shared memory architectures but with markedly different hardware capabilities to manage parallelism.

Triadic analysis, triads, triad census, large-scal↗

Shared Memory Parallelization of an Implicit ADI-type CFD Code

A parallelization study designed for ADI-type algorithms is presented using the OpenMP specification for shared-memory multiprocessor programming. Details of optimizations specifically addressed to cache-based computer architectures are described and performance measurements for the single and multiprocessor implementation are summarized. The paper demonstrates that optimization of memory access on a cache-based computer architecture controls the performance of the computational algorithm. A hybrid MPI/OpenMP approach is proposed for clusters of shared memory machines to further enhance the parallel performance. The method is applied to develop a new LES/DNS code, named LESTool. A preliminary DNS calculation of a fully developed channel flow at a Reynolds number of 180, Re(sub tau) = 180, has shown good agreement with existing data.

Hauser, Th.↗

Network-Attached Solid-State Recorder Architecture

A document discusses placing memory modules on the high-speed serial interconnect, which is used by a spacecraft s computer elements for inter-processor communications, to allow all multiple computer system architectures to access the spacecraft data storage at the same time. Each memory board is identical electrically and receives its bus ID upon connection to the system. The computer elements are configured in a similar fashion. The architecture allows for multiple memory boards to be accessed simultaneously by different computer elements, and results in a scalable, strong, fault-tolerant system. The IEEE-1393 ring bus can be routed so that multiple card failures can occur and the mass memory storage will still function.

Cox, Brian↗

Collective neural network behavior in a dynamically driven disordered system of superconducting loops

Collective properties of complex systems composed of many interacting components such as neurons in our brain can be modeled by artificial networks based on disordered systems. We show that a disordered neural network of superconducting loops with Josephson junctions can exhibit computational properties like categorization and associative memory in the time evolution of its state in response to information from external excitations. Superconducting loops can trap multiples of fluxons in many discrete memory configurations defined by the local free energy minima in the configuration space of all possible states. A memory state can be updated by exciting the Josephson junctions to fire or allow the movement of fluxons through the network as the current through them surpasses their critical current thresholds. Simulations performed with a lumped element circuit model of a 4-loop network show that information written through excitations is translated into stable states of trapped flux and their time evolution. Experimental implementation on a high-Tc superconductor YBCO-based 4-loop network shows dynamically stable flux flow in each pathway characterized by the correlations between junction firing statistics. Neural network behavior is observed as energy barriers separating state categories in simulations in response to multiple excitations, and experimentally as junction responses characterizing different flux flow patterns in the network. The state categories that produce these patterns have different temporal stabilities relative to each other and the excitations. This provides strong evidence for time-dependent (short-to-long-term) memories, that are dependent on the geometrical and junction parameters of the loops, as described with a network model.

Josephson junctions↗

Cache coherence for processing in memory

A cache coherence bridge protocol provides an interface between a cache coherence protocol of a host processor and a cache coherence protocol of a processor-in-memory, thereby decoupling coherence mechanisms of the host processor and the processor-in-memory. The cache coherence bridge protocol requires limited change to existing host processor cache coherence protocols. The cache coherence bridge protocol may be used to facilitate interoperability between host processors and processor-in-memory devices designed by different vendors and both the host processors and processor-in-memory devices may implement coherence techniques among computing units within each processor. The cache coherence bridge protocol may support different granularity of cache coherence permissions than those used by cache coherence protocols of a host processor and/or a processor-in-memory. The cache coherence bridge protocol uses a shadow directory that maintains status information indicating an aggregate view of copies of data cached in a system external to a processor-in-memory containing that data.

97 MATHEMATICS AND COMPUTING↗

Accelerating DCA++ (Dynamical Cluster Approximation) Scientific Application on the Summit supercomputer

Optimizing scientific applications on today’s accelerator-based high performance computing systems can be challenging, especially when multiple GPUs and CPUs with heterogeneous memories and persistent non-volatile memories are present. An example is Summit, an accelerator-based system at the Oak Ridge Leadership Computing Facility (OLCF) that is rated as the world’s fastest supercomputer to-date. New strategies are thus needed to expose the parallelism in legacy applications, while being amenable to efficient mapping to the underlying architecture.In this paper we discuss our experiences and strategies to port a scientific application, DCA++, to Summit. DCA++ is a highperformance research application that solves quantum manybody problems with a cutting edge quantum cluster algorithm, the dynamical cluster approximation.Our strategies aim to synergize the strengths of the different programming models in the code. These include: (a) streamlining the interactions between the CPU threads and the GPUs, (b) implementing computing kernels on the GPUs and decreasing CPU-GPU memory transfers, (c) allowing asynchronous GPU communications, and (d) increasing compute intensity by combining linear algebraic operations.Full-scale production runs using all 4600 Summit nodes attained a peak performance of 73.5 PFLOPS with a mixed precision implementation.We observed a perfect strong and weak scaling for the quantum Monte Carlo solver in DCA++, while encountering about 2× input/output (I/O) and MPI communication overhead on the time-to-solution for the full machine run. Our hardware agnostic optimizations are designed to alleviate the communication and I/O challenges observed, while improving the compute intensity and obtaining optimal performance on a complex, hybrid architecture like Summit.

Balduzzi, Giovanni↗

Sparse distributed memory

Theoretical models of the human brain and proposed neural-network computers are developed analytically. Chapters are devoted to the mathematical foundations, background material from computer science, the theory of idealized neurons, neurons as address decoders, and the search of memory for the best match. Consideration is given to sparse memory, distributed storage, the storage and retrieval of sequences, the construction of distributed memory, and the organization of an autonomous learning system.

Kanerva, Pentti↗

CRCNS22 Learning Rules in the Hippocampus and their Mapping to Neuromorphic Systems (Final Technical Report)

Large scale biologically-realistic computational models are key to investigating the interplay between structure and function in nervous systems, thus paving the way to new clinical methods and neuro-inspired computing solutions. This project focuses on the hippocampus, in particular the CA3-CA1 regions, due to their role in associative learning and memory, pattern separation and completion, and spatial navigation. Investigations into the neuronal organization and learning rule(s) of this circuit can shed light into how declarative memories are formed, stored, recalled and forgotten and inform computational, experimental and clinical neuroscience work. Our project aims at developing a novel data-driven methodology supported by a broad heterogeneous base of neuroscience experimental knowledge and inspired from advances in computer science and engineering. Specifically, this work will benchmark existing and new learning rules within a full-scale spiking neural network simulation of the CA3-CA1 region. The model will be based on an open-source repository, called the Hippocampome, which contains neuronal morphologies, firing patterns, synapse probabilities, and most other required parameters for all known neuron types in the rodent hippocampal formation. The model will be first trained in a supervised fashion for associative memory tasks using backpropagation through time traditionally used in computer science, enhanced with a new technique called the surrogate gradient method. This optimization method will be used to obtain a global loss minimization, but it is not biologically inspired as it assumes the use of data not locally available to the synapses. However, we propose its use as a benchmarking tool, to compare the training performance of local biologically plausible and hardware-mappable learning rules at scale. New rules or combinations will be proposed and tested as needed, based on the obtained results. Progress in this area will also drive the development of novel hardware-mappable algorithms for continual lifelong learning and categorization of new events from few presented examples. This project goes beyond the existing state-of-the-art by looking at large scale realistic neuronal circuits as networks trainable via global optimization methods such as surrogate gradient descent. The objective function of the brain that supports learning is largely unknown, but it is likely that it operates through local learning rules. Studying network trajectories around local minima as proposed in this work represents a useful strategy for understanding whether a network is training by using a specific (set of) learning rule(s). Starting from a completely untrained network is a challenging test since it is difficult to determine how the learning rule affects the trajectory of the network. This interdisciplinary project will help understand what rule governs learning in these regions or if multiple learning rules are involved. The work will develop a robust methodology to measure if the network is converging to the target solution, oscillating around it, or diverging away.

59 BASIC BIOLOGICAL SCIENCES↗

FastTree: Computing Large Minimum Evolution Trees with Profiles instead of a Distance Matrix

Gene families are growing rapidly, but standard methods for inferring phylogenies do not scale to alignments with over 10,000 sequences. We present FastTree, a method for constructing large phylogenies and for estimating their reliability. Instead of storing a distance matrix, FastTree stores sequence profiles of internal nodes in the tree. FastTree uses these profiles to implement Neighbor-Joining and uses heuristics to quickly identify candidate joins. FastTree then uses nearest neighbor interchanges to reduce the length of the tree. For an alignment with N sequences, L sites, and a different characters, a distance matrix requires O(N2) space and O(N2L) time, but FastTree requires just O(NLa+N √N) memory and O(N √N log(N)La) time. To estimate the tree’s reliability, FastTree uses local bootstrapping, which gives another 100-fold speedup over a distance matrix. For example, FastTree computed a tree and support values for 158,022 distinct 16S ribosomal RNAs in 17 h and 2.4 GB of memory. Just computing pairwise Jukes–Cantor distances and storing them, without inferring a tree or bootstrapping, would require 17 h and 50 GB of memory. In simulations, FastTree was slightly more accurate than Neighbor-Joining, BIONJ, or FastME; on genuine alignments, FastTree’s topologies had higher likelihoods. FastTree is available at http://microbesonline.org/fasttree.

59 BASIC BIOLOGICAL SCIENCES↗

Field-Deployable Quantum Memory for Quantum Networking

High-performance quantum memories are an essential component for regulating temporal events in quantum networks. As a component in quantum-repeaters, they have the potential to support the distribution of entanglement beyond the physical limitations of fiber loss. This will enable key applications such as quantum key distribution, network-enhanced quantum sensing, and distributed quantum computing. Here, we present a quantum memory engineered to meet real-world deployment and scaling challenges. The memory technology utilizes a warm rubidium vapor as the storage medium, and operates at room temperature, without the need for vacuum- and/or cryogenic- support. Here, we demonstrate performance specifications of high-fidelity retrieval (95%) and low operation error (10 –2 ) at a storage time of 160 µs for single-photon level quantum memory operations. We further show a substantially improved storage time (with classical-level light) of up to 1ms by suppressing atomic diffusions. The device is housed in an enclosure with a standard 2U rackmount form factor, and can robustly operate on a day scale in a noisy environment. This result marks an important step toward implementing quantum networks in the field.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Reduced scaling extended multi-state CASPT2 (XMS-CASPT2) using supporting subspaces and tensor hyper-contraction

We present a reduced scaling formulation of the extended multi-state CASPT2 (XMS-CASPT2) method, which is based on our recently developed state-specific CASPT2 (SS-CASPT2) formulation using supporting subspaces and tensor hyper-contraction. By using these two techniques, the off-diagonal elements of the effective Hamiltonian can be computed with only O(N 3 ) operations and O(N 2 ) memory, where N is the number of basis functions. Furthermore, this limits the overall computational scaling to O(N 4 ) operations and O(N 2 ) memory. Thus, excited states can now be obtained at the same reduced (relative to previous algorithms) scaling we achieved for SS-CASPT2.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Two-dimensional Euler and Navier-Stokes Time accurate simulations of fan rotor flows

Two numerical methods are presented which describe the unsteady flow field in the blade-to-blade plane of an axial fan rotor. These methods solve the compressible, time-dependent, Euler and the compressible, turbulent, time-dependent, Navier-Stokes conservation equations for mass, momentum, and energy. The Navier-Stokes equations are written in Favre-averaged form and are closed with an approximate two-equation turbulence model with low Reynolds number and compressibility effects included. The unsteady aerodynamic component is obtained by superposing inflow or outflow unsteadiness to the steady conditions through time-dependent boundary conditions. The integration in space is performed by using a finite volume scheme, and the integration in time is performed by using k-stage Runge-Kutta schemes, k = 2,5. The numerical integration algorithm allows the reduction of the computational cost of an unsteady simulation involving high frequency disturbances in both CPU time and memory requirements. Less than 200 sec of CPU time are required to advance the Euler equations in a computational grid made up of about 2000 grid during 10,000 time steps on a CRAY Y-MP computer, with a required memory of less than 0.3 megawords.

Boretti, A. A.↗

Memory protection

Accidental overwriting of files or of memory regions belonging to other programs, browsing of personal files by superusers, Trojan horses, and viruses are examples of breakdowns in workstations and personal computers that would be significantly reduced by memory protection. Memory protection is the capability of an operating system and supporting hardware to delimit segments of memory, to control whether segments can be read from or written into, and to confine accesses of a program to its segments alone. The absence of memory protection in many operating systems today is the result of a bias toward a narrow definition of performance as maximum instruction-execution rate. A broader definition, including the time to get the job done, makes clear that cost of recovery from memory interference errors reduces expected performance. The mechanisms of memory protection are well understood, powerful, efficient, and elegant. They add to performance in the broad sense without reducing instruction execution rate.

Denning, Peter J.↗

Optoelectronic Terminal-Attractor-Based Associative Memory

Report presents theoretical and experimental study of optically and electronically addressable optical implementation of artificial neural network that performs associative recall. Shows by computer simulation that terminal-attractor-based associative memory can have perfect convergence in associative retrieval and increased storage capacity. Spurious states reduced by exploiting terminal attractors.

Liu, Hua-Kuang↗