[Date Index] [Thread Index] [Author Index]

Re: Timing and Pi

• To: mathgroup at smc.vnet.net
• Subject: [mg84314] Re: Timing and Pi
• From: David Bailey <dave at Remove_Thisdbailey.co.uk>
• Date: Wed, 19 Dec 2007 04:07:56 -0500 (EST)
• References: <fk88qk$501$1@smc.vnet.net>

peter wrote:
> i wish to see other people results in running the code:
>   Timing[N[Pi,1000000]]
>   i am using mathematica 6 ,windowsXp
>   i have tried the code on two machines:
>   1- celeron 2.6GHZ-256kb cash, memory DDR-512MB-400Mhz
>   the timing result is  4.281 seconds
>   2- pentium D-925 dual processor 3GHZ-4MB cash , memory DDR2-2GB-667Mhz
>   the timing is 3.218 seconds
>   i am astonished that there is not much difference between the two values, even the second machine are speedier in cpu frequency and have more cash and the memory frequency.
>   thanks for all
>   peter

Don't forget that the Pi calculation will not use the second processor, 
and may not reference a lot of memory while it is doing the actual 
calculation - so the extra cache may not contribute too much either!

David Bailey
http://www.dbaileyconsultancy.co.uk