Parallel-processing a large scientific problem

Abstract

We discuss a parallel-processing experiment that uses a particle-in-cell (PIC) c(,[le to study the feasibility of doing large-scale scientific calculations on mult:ple-processor architectures. A multithread version of Lhis Los A]amos PIC code was successfully implemented and timed on a UNIVAC. System 1100/80 computer, Use of a single copy of the Instruction stream, and common memory to hold data, rlimin.ll("d dalii transmission between processors. "I"hr mul[ iplv-processing algorithm exploit% the' PIC~*ode's high dcgrrc of Iarg(', independent tasks, ;Js w(!11 IS Lhl' configuration of Lhe UNIVAC Syst.crn l'10()/lJO. Timing results lor Ihc multithrcad vu]sion [)! the PIC C(MI(' using onr, t,W(], thrwr. and four idrnt iral procr~:; ors arc given and arc shown to h.~vc promising sprvdIIp t imes whrn rompilrcd 10 the ovrra] 1 run t imrs mr;lsurcd for a singir-thread version of thr lJIc coctr.