Thread-Level Parallelism - Computer Architecture: A Quantitative Approach - page 408

linear, but so is the cost, since unlike TPC-C the amount of main memory and disk required
both scale less than linearly.
FIGURE 5.32 Speedup for three benchmarks on an IBM eServer p5 multiprocessor
when configured with 4, 8, 16, 32, and 64 processors . The dashed line shows linear spee-
dup.
As Figure 5.33 shows, larger processor counts can actually be more cost effective than the
four-processor configuration. In comparing the cost-performance of two computers, we must
be sure to include accurate assessments of both total system cost and what performance is
achievable. For many applications with larger memory demands, such a comparison can dra-
matically increase the atractiveness of using a multiprocessor.