Very slow performance on faster machine

John Hinsley jhinsley at telinco.co.uk
Wed Jun 27 05:42:14 UTC 2001


Tim Rowledge wrote:
> 
> John Hinsley <jhinsley at telinco.co.uk> is widely believed to have written:
> 
> > The only way to tell is to install some pretty sophisticated
> > benchmarking stuff or, easy peasy, take one block of memory out at a
> > time and see if 1 factorial runs quicker. Incidentally, on my 333, 128mb
> > on Linux, 100 factorial takes about 3 seconds.
> That seems oddly slow for what should be a fast machine:- my Acorn
> (200MHz cpu, negligable (like, 32Kb) cache, slow memory, five year old C
> compiler) does
>   Time millisecondsToRun:[Transcript show: 100 factorial printString]
> in 220 mS. Without the Transcript show: it is 100mS.
>   Time millisecondsToRun:[Transcript show: 1 factorial printString]
> takes 90mS and without the Transcript show:, <1mS.

Ah, I was trying to time it manually! (What an idiot!) Using your method
I get:

199 with Transcript show:

3 without it.

But i'm still very impressed by the performance of your Acorn!

I just wonder, before Tim tears that box down, whether it needs a good
defragging? I've heard of NT boxes coming to a grinding halt without it.

Time for bed.

Cheers

John
> 
> tim
> 
> --
> Tim Rowledge, tim at sumeru.stanford.edu, http://sumeru.stanford.edu/tim
> Strange OpCodes: MD: Move and Drop bits

-- 
******************************************************************************
Marx: "Why do Anarchists only drink herbal tea?"
Proudhon: "Because all proper tea is theft."
******************************************************************************





More information about the Squeak-dev mailing list