The performance evaluations of SSS-PC are provided.
These graphs are execution results of a parallelized version of POV-Ray for SSS-PC. Speedup ratio to 1-node execution is shown. Larger ratio value indicates faster execution. These values are evaluated from POV-Ray trace calculation time. Calculation time on tracing nodes and communication to displaying node are included in execution time. Parsing time (usually several seconds) on tracing nodes and output time on displaying node are not included. Xeon1, Xeon2, Xeon4, Xeon8 and Xeon16 are results of runs on 1, 2, 4, 8, 16 Pentium4-Xeon PCs respectively. pIII1, pIII2, pIII4 and pIII8 are results of runs on 1, 2, 4, 8, 16 PentiumIII PCs respectively. Four results for abyss.pov, chess2.pov, isocacti.pov and landscape.pov are shown.