S. Leung and J. Zahorjan. Extending the domain and improving the execution performance of runtime parallelization. Technical Report , in preparation, Department of Computer Science & Engineering, University of Washington, October 1992.  Home/Search   Document Not in Database   Summary   Related Articles

....shuntak cs.washington.edu, zahorjan cs.washington.edu. Because in this paper we are optimizing inspector time, but leaving the executor unchanged, the techniques we present have most dramatic effect when the inspector must be run for each invocation of the source loop. In a companion paper [3], we explore techniques that build upon those developed here to also improve executor performance. 1 Introduction Parallelizing compilers offer the parallel machine programmer the best of both worlds: the programming model is nearly identical to that of the familiar sequential programming ....

....The executor shown in Figure 3, first proposed by Saltz et al. 6, 8] and adopted for all work presented here, is sufficient to do this for the class of source loops under consideration. Improvements to both executor efficiency and domain of applicability are developed in a companion paper [3]. do w = 1 to depth pardo all i such that wf[i] w if (g(i) i) then a1 = anew[g(i) else a1 = aold[g(i) endif . anew[i] F(a1, a2, enddo enddo aold = anew Figure 3: Basic Executor Structure As written in Figure 2, the inspector is sequential. Because it processes as many iterations as ....

[Article contains additional citation context not shown here]

S. Leung and J. Zahorjan. Extending the domain and improving the execution performance of runtime parallelization. Technical Report , in preparation, Department of Computer Science & Engineering, University of Washington, October 1992.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC