p4 is a portable library of C and Fortran subroutines for programming parallel computers. It is the current version of a system that has been in use since 1984. It includes features for explicit parallel programming of shared-memory machines, distributed-memory machines (including heterogeneous networks of workstations), and clusters, by which we mean sharedmemory multiprocessors communicating via message passing. We discuss here the design goals, history, and system architecture of p4 and describe briefly a diverse collection of applications that have demonstrated the utility of p4. 1 Introduction p4 is a library of routines designed to express a wide variety of parallel algorithms portably, efficiently and simply. The goal of portability requires it to use widely accepted models of computation rather than specific vendor implementations of those models. The goal of efficiency requires it to use models of computation relatively close to those provided by the machines themselves and t...

Abstract. Parallel tools rely on graphical techniques to improve the quality of user interaction. In this paper, we explore how visualization and direct manipulation can be exploited in parallel tools, in order to improve the naturalness with which the user interacts with a parallel tool. Examples from recent tool research demonstrate that tool displays can be made more communicative and more intuitive touse. Visualization methods can be used to organize complex performance data into layers and perspectives that exploit the user's visual searching capabilities. Direct manipulation techniques allow the user to focus on key elements and then transition smoothly to further levels of detail or interrelated aspects of program behavior. Heuristics derived from studies with parallel users are proposed for when and how the techniques can be applied more e ectively. 1

Fine tuning the performance of large parallel programs is a very difficult task. A profiling tool can provide detailed insight into the utilization and communication of the different processors, which helps identify performance bottlenecks. In this paper we present a profiler for the fine-grained parallel programming language Split-C, which provides a simple global address space memory model. As our experience shows, it is much more challenging to profile programs that make use of efficient, low-overhead communication. We incorporated techniques which minimize profiling effects on the running program. We quantify the profiling overhead and present several Split-C applications which show that the profiler is useful in determining performance bottlenecks.