RE: Re: split a list
- To: mathgroup at smc.vnet.net
- Subject: [mg40643] RE: [mg40639] Re: split a list
- From: "Wolf, Hartmut" <Hartmut.Wolf at t-systems.com>
- Date: Sat, 12 Apr 2003 03:08:09 -0400 (EDT)
- Sender: owner-wri-mathgroup at wolfram.com
>-----Original Message----- >From: Bill Rowe [mailto:listuser at earthlink.net] To: mathgroup at smc.vnet.net >Sent: Friday, April 11, 2003 8:06 AM >To: mathgroup at smc.vnet.net >Subject: [mg40643] [mg40639] Re: split a list > > >On 4/10/03 at 3:44 AM, majort at cox-internet.com (Dr Bob) wrote: > >>Your solution (the same as Jens-Peer Kuska's) IS simpler and clearer >>(though clarity is in the eye of the beholder!) > >Agreed > >>-- but my timing comparisons disagree with yours. Perhaps my >>environment is different (WinXP, 1024MB Ram, 2.2 GHz P4, Mathematica >>4.2.1). > >Yes, my enviroment is different (MacOS X, 1024MB Ram, 800 MHz >G4, Mathematica 4.2.1). With a different environment it isn't >surprising the actual times are different. But I would have >thought the time rankings would be the same given the same >version of Mathematica. It seems interesting this doesn't seem >to be the case. Clearly a case of user beware. > >In any case, the timing differences between the two methods is >quite small and unlikely to be noticed except for extreme cases. > Bill, this is not a surprising phenomenon. As the total computation time is not the result of a single hardware parameter (say CPU clock), but involves quite a row: instruction set, CPU architecture, management of the instruction pipeline, caching hierachy, strategy and cache access times and finally core memory access. (You may possibly change some paramters with BIOS set-up.) Also, quite often, machine code is optimized in different ways for different machine architectures. This is why for comparing machines, normally a whole suite of benchmarks is used. -- Hartmut Wolf