关键词： 多处理器   实时系统   周期任务   容错调度  

摘要： 主副版本策略是多处理器系统实时任务调度中处理容错问题的一种重要方式.根据分布式控制系统的特点,本文提出一种改进的FTRMBF算法—PR-FTRMBF,以提高系统周期任务的可调度性.在FTRMBF等已有的调度算法中,当没有处理器分配给当前副版本时,将为副版本分配新的处理器;本文提出的改进算法则以回溯的方式重新分配主版本.在保证系统实时性能和容错能力的前提下,节省了处理器数目.仿真实验表明,与FTRMBF算法相比,改进算法显著提高了系统任务的可调度性.

关键词： 并行计算   应用   多处理器   美国能源部   超级计算机   大型工程   复杂网络   并行算法  

摘要： 美国能源部将“并行计算”定义为：在一个应用中用多个处理器同时来处理，它是大型科学、大型工程和大量商业计算的根本途径。本书为计算科学家和工程师提供了一个实用的指导，帮助他们利用具有多处理器与复杂网络的超级计算机来促进自己的研究。本书的重点是基本并行算法的设计和分析。

关键词： 异构多处理器   任务调度   PSO算法   模拟退火   遗传算法  

摘要： 为提高异构多处理器任务调度的执行效率,充分发挥多处理器并行性能,提出一种基于粒子群优化的异构多处理器任务调度算法-PSOASA算法.PSOASA算法以求得任务最短完成时间为目标,首先采用整数矩阵对粒子进行编码,并定义交换操作更新粒子状态,实现粒子搜索空间到离散空间的映射,使连续的粒子群优化算法适用于离散的异构多处理器任务调度问题,同时引入模拟退火算法,克服粒子群算法的"早熟"收敛现象,避免求得的解陷入局部最优.实验结果表明,PSOASA算法的执行效率优于目前广泛采用的遗传算法,有效地降低任务执行时间,减少了迭代次数,适用于异构多处理器环境大规模任务调度.

关键词： OpenMP   多处理器   高斯-约当消元法   并行处理  

摘要： 文章介绍了OpenMP的并行执行原理和语言规范,讨论了OpenMP的循环并行化、迭代相关、数据共享、任务调度等问题。接着研究了高斯-约当消元法固有的并行性,提出并行高斯-约当消元法,并基于多处理器平台HPZ620进行了测试。实验结果表明,理论分析与实验结果是一致的。

关键词： Heterogeneous (hybrid) systems   distributed architecture   natural language interfaces   machine learning  

摘要： Given the recent progress in the evolution of high-performance computing (HPC) technologies, the research in computational intelligence has entered a new era. In this paper, we present an HPC-based context-aware intelligent text recognition system (ITRS) that serves as the physical layer of machine reading. A parallel computing architecture is adopted that incorporates the HPC technologies with advances in neuromorphic computing models. The algorithm learns from what has been read and, based on the obtained knowledge, it forms anticipations of the word and sentence level context. The information processing flow of the ITRS imitates the function of the neocortex system. It incorporates large number of simple pattern detection modules with advanced information association layer to achieve perception and recognition. Such architecture provides robust performance to images with large noise. The implemented ITRS software is able to process about 16 to 20 scanned pages per second on the 500 trillion floating point operations per second (TFLOPS) Air Force Research Laboratory (AFRL)/Information Directorate (RI) Condor HPC after performance optimization.

关键词： Topology  

摘要： On-chip network is a promising solution for the on-chip communication problem of large-scale CMPs and a major factor in the performance, area, and power consumption of the overall system. This paper proposes an on-chip network with regular reconfigurable topology (RROCN). The RROCN is a reconfigurable and hybrid communication structure contained routed network with 2D mesh topology and shared bus. The reconfiguration is implemented by disabling and bypassing the unwanted nodes of routed network and then organizing them as shared buses. To achieve this goal, a constructive algorithm, reconfiguration scheme, and modified XY routing algorithm with self-adaptive feature are proposed. We evaluate the RROCNs with four reconfiguration topologies and compare them with a regular on-chip network. Our results show that, with 4×4 2D mesh reconfigurable topology, the RROCN requires 41.3% less power, provides 34.5% lower zero-load latency, but provides 34.6% lower maximum throughput compared with the regular on-chip network.

关键词： Sensor network   reliability   dynamic control   low-power   noise   power grid   system on chip  

摘要： Reducing feature sizes and power supply voltage allows integrating more processing units (PUs) on multiprocessor system on chip (MPSoC) to satisfy the increasing demands of applications. However, it also makes MPSoC more susceptible to various reliability threats, such as high temperature and power/ground (P/G) noise. As the scale and complexity of MPSoC continuously increase, monitoring and mitigating reliability threats at runtime could offer better performance, scalability, and flexibility for MPSoC designs. In this paper, we propose a systematic approach, on-chip sensor network (SENoC), to collaboratively predict, detect, report, and alleviate runtime threats in MPSoC. SENoC not only detects reliability threats and shares related information among PUs, but also plans and coordinates the reactions of related PUs in MPSoC. SENoC is used to alleviate the impacts of simultaneous switching noise in MPSoC's P/G network during power gating. Based on the detailed noise behaviors under different scenarios derived by our circuit-level MPSoC P/G noise simulation and analysis platform, simulation results show that SENoC helps to achieve on average 26.2 percent performance improvement compared with the traditional stop-go method with 1.4 percent area overhead in an 8*8-core MPSoC in 45 nm. An architecture-level cycle-accurate simulator based on SystemC is implemented to study the performance of the proposed SENoC. By applying sophisticated scheduling techniques to optimize the total system performance, a higher performance improvement of 43.5 percent is achieved for a set of real-life applications.

关键词： 多处理器   任务调度   负载均衡   细菌觅食优化算法  

摘要： 为了提高多核处理器任务调度的效率和负载均衡,提出一种改进细菌觅食优化算法的多核处理器任务调度方法.首先随机产生多核处理器任务有向无环图,然后构建一种多约束条件下的数学模型,最后采用改进细菌觅食优化算法对模型进行求解.仿真结果表明,改进细菌觅食优化算法很好避免了其它群智算法存在的缺陷,可以较好获得全局最优调度方案,使执行任务的时间最短,很好地实现了多核处理器任务进行的分配和调度.

关键词： 智能相机   BOA   工业自动化技术   参展产品   技术服务   机器视觉   多处理器   外观设计  

摘要： 2013年3月4日-6日，作为从事机器视觉产品的推广及技术服务的专业公司，上海波创携最新产品亮相广州国际工业自动化技术及装备展。在国际自动化新品云集的展会上，上海波创的新一代多处理器智能相机BOA，以其强大的性能及流线的外观设计成为同类参展产品中耀眼的明星。

关键词： Scheduling  

摘要： Multicore and multiprocessor systems with dynamic voltage scaling architectures are being used as one of the solutions to satisfy the growing needs of high performance applications with low power constraints. An important aspect that has propelled this solution is effective task/application scheduling and mapping algorithms for multiprocessor systems. This work proposes an energy aware, offline, probability-based unified scheduling and mapping algorithm for multiprocessor systems, to minimize the number of processors used, maximize the utilization of the processors, and optimize the energy consumption of the multiprocessor system. The proposed algorithm is implemented, simulated and evaluated with synthetic task graphs, and compared with classical scheduling algorithms for the number of processors required, utilization of processors, and energy consumed by the processors for execution of the application task graphs. © 2013. The Korean Institute of Information Scientists and Engineers.