57 (b) Cutting the disk I/O time will substantially improve the external sorting time. A reasonable estimate is that cutting disk I/O time in half will cut the sorting time by around 1/3. (c) Main memory access time will not help a great deal, since disk I/O is the probable bottleneck. However, for the sorting operation, main memory access time is in fact more of a bottleneck than CPU speed, so it should help more to speed the memory than to speed the CPU. (d) Increasing the memory size by a factor of two will increase the f le size that can be processed by a single pass of multi-way merge by a factor of four, in two passes by a factor of eight, and so no. If this leads to a reduction in the number of passes need to process the f le, then a substantial time savings will be realized. This could easily cut the processing time by 1/3 or 1/4 since 2 or 3 passes of multiway merge under the initial conditions are reasonable to expect. 8.14
This is the end of the preview. Sign up
access the rest of the document.