Chris, Thanks for the correction. Yes, Log2 would offer a better guide than logN for the analysis of sort performance characteristics.
John P. Baker -----Original Message----- From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of Blaicher, Chris Sent: 03/10/2008 10:34 AM To: [email protected] Subject: Re: Exponential growth (was how fast can I sort on mainframe (using DFSORT)?) In this context, sorting and the exponential nature of it, it would have been better to use a log base 2 example, rather than a natural log example. The generalized cost of sorting is log2(n) compares per record. That said, in the last 15 years CPU time has not been a major factor in how long a sort runs. It has been the data transfer time that is the limiting factor on sort elapsed time. 99% of the CPU time is overlapped with I/O time. CPU time is a resource and should not be wasted, no question about that, but it is probably not a factor in the original poster's question. Chris Blaicher ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
