SEARCH · Search NASA
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.
The paper describes the architectual details of a fine-grained Programmable Transistor Array (PTA) architecture and illustrates its use in evolutionary experiments on the synthesis of both analog and digital circuits.
In this paper a new algorithm, designated as Fast Invariant Imbedding algorithm, for solution of Poisson equation on vector and massively parallel MIMD architectures is presented. This algorithm achieves the same optimal computational efficiency as other Fast Poisson solvers while offering a much better structure for vector and parallel implementation. Our implementation on the Intel Delta and Paragon shows that a speedup of over two orders of magnitude can be achieved even for moderate size problems.
The performance modeling and enhancement for periodic execution of large-grain, decision-free algorithms in data flow architectures is examined. Applications include real-time implementation of control and signal processing algorithms where performance is required to be highly predictable. The mapping of algorithms onto the specified class of data flow architectures is realized by a marked graph model called ATAMM (Algorithm To Architecture Mapping Model). Performance measures and bounds are established. Algorithm transformation techniques are identified for performance enhancement and reduction of resource (computing element) requirements. A systematic design procedure is described for generating operating conditions for predictable performance both with and without resource constraints. An ATAMM simulator is used to test and validate the performance prediction by the design procedure. Experiments on a three resource testbed provide verification of the ATAMM model and the design procedure.
Performance modeling and performance enhancement for periodic execution of large-grain, decision-free algorithms in data flow architectures are discussed. Applications include real-time implementation of control and signal processing algorithms where performance is required to be highly predictable. The mapping of algorithms onto the specified class of data flow architectures is realized by a marked graph model called algorithm to architecture mapping model (ATAMM). Performance measures and bounds are established. Algorithm transformation techniques are identified for performance enhancement and reduction of resource (computing element) requirements. A systematic design procedure is described for generating operating conditions for predictable performance both with and without resource constraints. An ATAMM simulator is used to test and validate the performance prediction by the design procedure. Experiments on a three resource testbed provide verification of the ATAMM model and the design procedure.
A functional description of the ATAMM Multicomputer Operating System is presented. ATAMM (Algorithm to Architecture Mapping Model) is a marked graph model which describes the implementation of large grained, decomposed algorithms on data flow architectures. AMOS, the ATAMM Multicomputer Operating System, is an operating system which implements the ATAMM rules. A first generation version of AMOS which was developed for the Advanced Development Module (ADM) is described. A second generation version of AMOS being developed for the Generic VHSIC Spaceborne Computer (GVSC) is also presented.
Spectral element methods are high-order weighted residual techniques for partial differential equations that combine the geometric flexibility of finite element methods with the rapid convergence of spectral techniques. Spectral element methods are described for the simulation of incompressible fluid flows, with special emphasis on implementation of spectral element techniques on medium-grained parallel processors. Two parallel architectures are considered: the first, a commercially available message-passing hypercube system; the second, a developmental reconfigurable architecture based on Geometry-Defining Processors. High parallel efficiency is obtained in hypercube spectral element computations, indicating that load balancing and communication issues can be successfully addressed by a high-order technique/medium-grained processor algorithm-architecture coupling.
Spectral element methods are high-order weighted residual techniques for partial differential equations that combine the geometric flexibility of finite element methods with the rapid convergence of spectral techniques. Spectral element methods are described for the simulation of incompressible fluid flows, with special emphasis on implementation of spectral element techniques on medium-grained parallel processors. Two parallel architectures are considered; the first, a commercially available message-passing hypercube system; the second, a developmental reconfigurable architecture based on Geometry-Defining Processors. High parallel efficiency is obtained in hypercube spectral element computations, indicating that load balancing and communication issues can be successfully addressed by a high-order technique/medium-grained processor algorithm-architecture coupling.
A software simulator capability of simulating execution of an algorithm graph on a given system under the Algorithm to Architecture Mapping Model (ATAMM) rules is presented. ATAMM is capable of modeling the execution of large-grained algorithms on distributed data flow architectures. Investigating the behavior and determining the performance of an ATAMM based system requires the aid of software tools. The ATAMM Simulator presented is capable of determining the performance of a system without having to build a hardware prototype. Case studies are performed on four algorithms to demonstrate the capabilities of the ATAMM Simulator. Simulated results are shown to be comparable to the experimental results of the Advanced Development Model System.
Diagnostics software for analyzing Algorithm to Architecture Mapping Model (ATAMM) based concurrent processing systems is presented. ATAMM is capable of modeling the execution of large grain algorithms on distributed data flow architectures. The tool graphically displays algorithm activities and processor activities for evaluation of the behavior and performance of an ATAMM based system. The tool's measurement capabilities indicate computing speed, throughput, concurrency, resource utilization, and overhead. Evaluations are performed on a simulated system using the software tool. The tool is used to estimate theoretical lower bound performance. Analysis results are shown to be comparable to the predictions.
The Navier-Stokes Computer (NSC) consists of multiple local memory parallel processors interconnected in a hypercube network. Efficient implementation of algorithms on the NSC thus requires the effective utilization of both the coarse and fine grain paralelism inherent in the architectural design. The basic approach to implementing an algorithm on the NSC is presented herein. The particular finite-difference algorithm considered was developed for performing transition and turbulence simulations by direct solution of the time-dependent incompressible Navier-Stokes equations. The suitability of this algorithm for performing simulations of the isotropic turbulence problem is verified from computations performed on a Cray 2. Projected timing results for the algorithm on the NSC itself are presented for both the isotropic turbulence and laminar turbulent transition problems.
Very Long Instruction Word (VLIW) architectures enhance performance by exploiting fine-grained instruction level parallelism. We describe the development of two compiler assisted multiple instruction word retry schemes for VLIW architectures. The first scheme utilizes the compiler techniques previously developed for processors with single functional units. A compiler generated hazard-free code with different degrees of rollback capability for uniprocessors is compacted by a modified VLIW trace scheduling algorithm. Nops are then inserted in the scheduled code words to resolve data hazards for VLIW architectures. Performance is compared under three parameters: the rollback distance for uni-processors; the number of functional units; and the rollback distance for VLIW architectures. The second scheme employs a hardware read buffer to resolve frequently occurring data hazards, and utilizes the compiler to resolve the remaining hazards. Performance results are shown for six benchmark programs.
In this paper the implementation of a parallel O(LogN) algorithm for computation of rigid multibody dynamics on a Hypercube MIMD parallel architecture is presented. To our knowledge, this is the first algorithm that achieves the time lower bound of O(LogN) by using an optimal number of O(N) processors. However, in addition to its theoretical significance, the algorithm is also highly efficient for practical implementation on commercially available MIMD parallel architectures due to its highly coarse grain size and simple communication and synchronization requirements. We present a multilevel parallel computation strategy for implementation of the algorithm on a Hypercube. This strategy allows the exploitation of parallelism at several computational levels as well as maximum overlapping of computation and communication to increase the performance of parallel computation.
The algorithm to architecture mapping model (ATAAM) is a Petri net based model which provides a strategy for periodic execution of a class of real-time algorithms on multicomputer dataflow architecture. The execution of large-grained, decision-free algorithms on homogeneous processing elements is studied. The ATAAM provides an analytical basis for calculating performance bounds on throughput characteristics. Extension of the ATAMM as a strategy for cyclo-static scheduling provides for a truly distributed ATAMM multicomputer operating system. An ATAAM testbed consisting of a centralized graph manager and three processors is described using embedded firmware on 68HC11 microcontrollers.
Computational Aero Sciences and other numeric intensive computation disciplines demand computing throughputs substantially greater than the Teraflops scale systems only now becoming available. The related fields of fluids, structures, thermal, combustion, and dynamic controls are among the interdisciplinary areas that in combination with sufficient resolution and advanced adaptive techniques may force performance requirements towards Petaflops. This will be especially true for compute intensive models such as Navier-Stokes are or when such system models are only part of a larger design optimization computation involving many design points. Yet recent experience with conventional MPP configurations comprising commodity processing and memory components has shown that larger scale frequently results in higher programming difficulty and lower system efficiency. While important advances in system software and algorithms techniques have had some impact on efficiency and programmability for certain classes of problems, in general it is unlikely that software alone will resolve the challenges to higher scalability. As in the past, future generations of high-end computers may require a combination of hardware architecture and system software advances to enable efficient operation at a Petaflops level. The NASA led HTMT project has engaged the talents of a broad interdisciplinary team to develop a new strategy in high-end system architecture to deliver petaflops scale computing in the 2004/5 timeframe. The Hybrid-Technology, MultiThreaded parallel computer architecture incorporates several advanced technologies in combination with an innovative dynamic adaptive scheduling mechanism to provide unprecedented performance and efficiency within practical constraints of cost, complexity, and power consumption. The emerging superconductor Rapid Single Flux Quantum electronics can operate at 100 GHz (the record is 770 GHz) and one percent of the power required by convention semiconductor logic. Wave Division Multiplexing optical communications can approach a peak per fiber bandwidth of 1 Tbps and the new Data Vortex network topology employing this technology can connect tens of thousands of ports providing a bi-section bandwidth on the order of a Petabyte per second with latencies well below 100 nanoseconds, even under heavy loads. Processor-in-Memory (PIM) technology combines logic and memory on the same chip exposing the internal bandwidth of the memory row buffers at low latency. And holographic storage photorefractive storage technologies provide high-density memory with access a thousand times faster than conventional disk technologies. Together these technologies enable a new class of shared memory system architecture with a peak performance in the range of a Petaflops but size and power requirements comparable to today's largest Teraflops scale systems. To achieve high-sustained performance, HTMT combines an advanced multithreading processor architecture with a memory-driven coarse-grained latency management strategy called "percolation", yielding high efficiency while reducing the much of the parallel programming burden. This paper will present the basic system architecture characteristics made possible through this series of advanced technologies and then give a detailed description of the new percolation approach to runtime latency management.
With the present availability of parallel processors of vastly different architectures, there is a need for a common language interface to multiple types of machines. The parallel C compiler, currently under development, is intended to be such a language. This language is based on the belief that an algorithm designed around fine-grained parallelism can be mapped relatively easily to different parallel architectures, since a large percentage of the parallelism has been identified. The compiler generates a FORTH-like machine-independent intermediate code. A machine-dependent translator will reside on each machine to generate the appropriate executable code, taking advantage of the particular architectures. The goal of this project is to allow a user to run the same program on such machines as the Massively Parallel Processor, the CRAY, the Connection Machine, and the CYBER 205 as well as serial machines such as VAXes, Macintoshes and Sun workstations.
A global study of megafans (greater than 100 km long) has revealed their widespread existence on all continents, with almost 200 documented, 93 in Africa where research is most thorough. The largest measures 705 km. Megafans are a major subset of "DFS" (distributive fluvial systems, a category that includes all fan-like features greater than 30 km long). 1. Many researchers now recognize megafans as different from floodplains, small coarse-grained alluvial fans, and deltas. Although smaller architectural elements in megafans are the same as those encountered in floodplains (channel, overbank, etc.), larger architectures differ because of the unconfined setting of megafans, versus the valley-confined setting of floodplains. 2. A length continuum is now documented between steep alluvial fans 10-20 km in length, and fluvial fans 30-50 km long. This implies a continuum of process from end-member alluvial fan processes (e.g. high-energy flows that emplace gravels, debris-flow units) to the relatively fine-grained channel and overbank deposits common to purely fluvial fans. Combinations of these different processes will then occur in many mid-sized fans. 3. The global distribution suggests a prima facie relationship with tectonic environment rather than climatic zones, with local controls being the slope of the formative river and the existence of a basin subsiding below the long profile of the river. But the global population has revealed that most megafans are relict. So it is possible that further research will show relationships to prior climatic regimes. 4. Megafans can have regional importance: e.g., along the east flank of the central Andes, nested megafans total approximately 750,000 km2-and 1.2m km2 if all megafans in S. America are counted. Modern megafan landscapes thus have basinal importance, orders of magnitude greater than alluvial fan bajadas. 5. Because so many aggrading basins are dominated today by DFS, it is claimed that DFS ought to be significant in the subsurface; and that existing fluvial models therefore may not apply to the majority of fluvial sedimentary units. Arguments have been raised against this view, but as modern megafan systems become better known they are rapidly being applied as a model in many fluvial basins. A small literature has arisen with apparent examples from every part of the world.
The analytical model for resource utilization and the variable node time and conditional node model for the enhanced ATAMM model for a real-time data flow architecture are presented in this research. The Algorithm To Architecture Mapping Model, ATAMM, is a Petri net based graph theoretic model developed at Old Dominion University, and is capable of modeling the execution of large-grained algorithms on a real-time data flow architecture. Using the resource utilization model, the resource envelope may be obtained directly from a given graph and, consequently, the maximum number of required resources may be evaluated. The node timing diagram for one iteration period may be obtained using the analytical resource envelope. The variable node time model, which describes the change in resource requirement for the execution of an algorithm under node time variation, is useful to expand the applicability of the ATAMM model to heterogeneous architectures. The model also describes a method of detecting the presence of resource limited mode and its subsequent prevention. Graphs with conditional nodes are shown to be reduced to equivalent graphs with time varying nodes and, subsequently, may be analyzed using the variable node time model to determine resource requirements. Case studies are performed on three graphs for the illustration of applicability of the analytical theories.
It is anticipated that in order to make effective use of many future high performance architectures, programs will have to exhibit at least a medium grained parallelism. A framework is presented for partitioning very sparse triangular systems of linear equations that is designed to produce favorable preformance results in a wide variety of parallel architectures. Efficient methods for solving these systems are of interest because: (1) they provide a useful model problem for use in exploring heuristics for the aggregation, mapping and scheduling of relatively fine grained computations whose data dependencies are specified by directed acrylic graphs, and (2) because such efficient methods can find direct application in the development of parallel algorithms for scientific computation. Simple expressions are derived that describe how to schedule computational work with varying degrees of granularity. The Encore Multimax was used as a hardware simulator to investigate the performance effects of using the partitioning techniques presented in shared memory architectures with varying relative synchronization costs.