Nowadays, most computers that are commercially available off-the-shelf (COTS) include hardware features that increase the performance of parallel general-purpose threads (hyper threading, multicore, ccNUMA architectur...
详细信息
We describe the MultiFlex multi-processor SoC programming environment, with the focus on two programming models: a distributed system object component (DSOC) message passing model, and a symmetrical multi-processing (...
详细信息
ISBN:
(纸本)9781581139372
We describe the MultiFlex multi-processor SoC programming environment, with the focus on two programming models: a distributed system object component (DSOC) message passing model, and a symmetrical multi-processing (SMP) model using shared memory. The MultiFlex tools map these models onto the StepNP multi-processor SoC platform, while making use of hardware accelerators for message passing and task scheduling. We present the results of mapping an Internet traffic management application, running at 2.5 Gb/s.
MMX (multiprocessor multitasking executive) is a small yet powerful operating system for shared memory multiprocessors. The MMX parallel processor is a small shared bus multiprocessor assembled from several commercial...
详细信息
MMX (multiprocessor multitasking executive) is a small yet powerful operating system for shared memory multiprocessors. The MMX parallel processor is a small shared bus multiprocessor assembled from several commercial processor boards. Together, MMX and its parallel processor provide a flexible and powerful testbed for parallel software development. The author describes MMX design principles, its structure and services. parallel programming techniques with MMX are demonstrated with special emphasis on work distribution. The author concludes with some MMX timing and speedup measurements of several parallel programs.< >
Several mesh-like architectures have been proposed to support parallel computing. These connect the processor mesh with a system of reconfigurable buses. In this work, the authors present several examples to illustrat...
详细信息
Several mesh-like architectures have been proposed to support parallel computing. These connect the processor mesh with a system of reconfigurable buses. In this work, the authors present several examples to illustrate the ease with which efficient (in fact, constant time) programs can be developed for the weakest of these architectures. The examples are from computational geometry.< >
Many different paradigms for parallel programming exist, nearly each of which is employed in dozens of languages. Several researchers have tried to compare these languages and paradigms by examining the expressivity a...
详细信息
This paper covers the design of a course introducing parallel programming. The design is partly based on the feedback collected by the author when teaching other parallel computing courses offered for senior undergrad...
详细信息
parallel applications with inconstant usage patterns presents a big challenge to programmers in that the spawning of tasks and the communication between them may be conditional (named »conditional parallel progra...
详细信息
parallelising serial software systems presents many challenges. In particular,the task of decomposing large, data-intensive applications for execution on distributedarchitectures is described in the literature as erro...
详细信息
parallelising serial software systems presents many challenges. In particular,the task of decomposing large, data-intensive applications for execution on distributedarchitectures is described in the literature as error-prone and *** Message Passing Interface (MPI) specification is the de facto industry standardto program for such architectures, but requires low level knowledge of datadistribution details as programmers must explicitly invoke inter-process communicationroutines. This research reports the findings from empirical studiesconducted in industry, to explore and characterise the challenges associated withperforming data decomposition. Findings from these studies culminated in alist of derived requirements for tool support, encompassing automation of gridindexing, generation of data structures and communication calls, and provisionof assistance when changing from an implemented decomposition strategy. Additionalrequirements include the need for a tool to be MPI focused, initiallytarget structured grids and have a low impact on the application code. Theserequirements were subsequently buttressed to address gaps in the state-of-the-artand provided motivation for the development of a tool named *** provides an abstraction for MPI, encapsulating the low level detailsinvolved in decomposing data and exchanging messages between *** can express the parallel intent of their application through input parametersand then generate code containing wrapper functions that encompass the MPIfunctionality. The wrapper functions can then be invoked within the serial coderesulting in a semi-automated parallelised solution. The programmer is relievedof the burden of deciphering memory locations when exchanging data betweenprocessors. The tool was evaluated in two studies involving both students andHigh Performance Computing (HPC) practitioners as subjects. The findingsconcluded that MPIGen provides an efficient abstraction for performing dat
The authors report how they used Ada's features for parallel processing, the task and the rendezvous, to achieve near-linear speedup on parallel computers that hosted an Ada parallel run-time environment. In addit...
详细信息
The authors report how they used Ada's features for parallel processing, the task and the rendezvous, to achieve near-linear speedup on parallel computers that hosted an Ada parallel run-time environment. In addition, they discuss the availability, performance, and productivity of Ada for the programmer familiar with single-program, multiple-data (SPMD)-style programming in parallel Fortran.< >
NASA Technical Reports Server (Ntrs) 19980237540: Integrated Task and Data parallel programming: Language Design by NASA Technical Reports Server (Ntrs); published by
NASA Technical Reports Server (Ntrs) 19980237540: Integrated Task and Data parallel programming: Language Design by NASA Technical Reports Server (Ntrs); published by
暂无评论