Speedup curves

Speedup curves illustrating commonly encountered performance patterns, (a) ideal (b) superlinear speedup with performance degradation due to, e.g., communication overhead or load imbalance (c) logarithmic communication overhead (d) linear communication overhead (e) incompletely parallelized program (serial fraction of 0.025). See text for details. 

Predicted and measured speedups and efficiencies for the simple parallel matrix-vector multiplication outlined in Figure 5.4. Dashed curves represent predictions by the performance model, solid curves show measured values, and the dot-dashed line is the ideal speedup curve. The... 

Speedups for MP2 algorithms PI and P2 (dynamic version) measured relative to timings for one process (Pli and P2i) and sixteen processes (Plie and P2i6). Computations tvere performed on a Linux cluster for the uracil dimer molecule using the cc-pVDZ basis set (cf. Figure 9.8). Inflated speedup curves are obtained by measuring speedups relative to a number of processes greater than one. 

A number of ways to report misleading parallel performance data have been discussed elsewhere/ including how to boost performance data by comparing with code that is nonoptimal in a number of ways. Performance data are most often presented in the form of speedup curves, and it is therefore important to ascertain that the presented speedups are, in fact, representative of the typical parallel performance of the algorithm. Below we will discuss a couple of commonly encoimtered practices for presenting speedups that can lead to misrepresentation of performance data. 

The difference in speedup curves for these two cases demonstrates that network performance characteristics can have significant impact on the parallel performance of an algorithm. 

Fig. 1. Amdahl s law. Speedup as a function of the percentage of the program that can be vectorized. Lower curve vector—scalar speedup = 10 upper curve...

Figure 6 Amdahl s law as a function of the number of processors. Each curve in this family of curves represents a different percentage of the code that runs in parallel. The Speedup(lOO) curve is the ideal curve because the code executes in parallel 100%.

---