IBM C24-3038-0 sort2timing free download

File name C24-3038-0_sort2timing.pdf

File Number GENt - 33 Form C24-3038-0 Systems Reference Library Comparison of Sort 2-Versi(Jn 2 Timing IBM 1401 Elnd IBM 1460 Data Processing Systems This pubLication contains 72 sort timing comparison tables of the Sort 2 - Version 2 program run on an IBM 1401 Data Processin~~ System and an IBM 1460 Data Processing System. This program, when used with a 1460 system, results in an appreciable savings in sort run tilne. The various parameters involved are expla.ined in the introduction of the publication. @ 1963 by International Business Machines Corporation COMPARISON OF SORT 2-VERSION 2 TIMING - IBM 1401 AND IBM 1460 DATA PROCESSING SYSTEMS With the announcement by IBM of the IBM 1460 Data Processing System, an appreciable savings in sort timing can be realized by the user. This publication contains sort timing comparison tables made by run- ning the Sort 2 - Version 2 program on an IBM 1401 Data Processing System and an IBM 1460 Data Pro- cessing System. The sort timing estimates are for 4-tape and 6 -tape sorts using the balanced Phase 2 method and 4-tape sorts using the multiphase method of mer- ging in Phase 2. The parameters involved are: * Core -storage capacity - 8, 000; 12, 000; and 16,000 positions * Magnetic tape units - 7330; 729-II; 729-IV; 729-V; and 729-VI (729-VI used with 1460 only) * Record size - 10, 20, 40, 80, 100, 200, 500, and 1000 characters * Input file size - 1000; 2000; 5000; 10,000; 25,000; 50,000; 75,000; and 100,000 records * Density - 556 characters per inch on 7330; 729-II; and 729-IV tape units. 800 characters per inch on 729-V and 729-VI tape units. The number of control field characters used for each table is 10 characters. COMPARISON TIMING TABLES This section contains 48 comparison timing tables for 2-way and 3-way balanced merges, and 24 tables for multiphase merging. All of the parameters pre- viously discussed have been used to compile these timing estimate tables. Also included are these three factors: G - the number of records sorted internally in Phase 1 B - the sort blocking factor P - the number of Phase 2 merge passes The timings shown have been rounded to the nearest full minute. In the tables with small record volume, the asterisk (*) indicates that this run was accomplished in less than one-half minute. Copies of this and other mM publications can be obtained through IBM Branch Offices Address comments regarding the content of this publication to IBM Product Publications, Endicott, New York