From: Kristof Bastiaensen <kristof@...>
Date: 2006-08-05T18:30:04+09:00
Subject: Re: For performance, write it in C - Part 3, Source code now available

On Sat, 05 Aug 2006 08:31:30 +0900, ara.t.howard wrote:

> On Sat, 5 Aug 2006, Isaac Gouy wrote:
> 
>>
>> ara.t.howard@noaa.gov wrote:
>>> On Sat, 5 Aug 2006, Peter Hickman wrote:
>>>
>>>> Perhaps stepping up to a 6 x 6 grid would allow more meaningful timings?
>>>
>>> showing how it performs on a 2, 4, and 8 grid would show how it scaled...
>>>
>>> -a
>>> --
>>> happiness is not something ready-made.  it comes from your own actions.
>>> - h.h. the 14th dali lama
>>
>> It scales badly
>>
>>        0.496s gcc 5x5 (without print statements)
>> 30055.098s gcc 6x6 (without print statements)
>>
>> imo If Peter has figured out what he's trying to do then it would be
>> smart to find a better way of doing it rather than scrabbling for small
>> percentage improvements.
> 
> indeed - that's my point exactly.
> 
> with that kind scaling a better ruby version could kill the c one!

Did you actually run the 6x6 version?  I think my Ruby version might
actually be faster than the C one for large squares, since it only
computes normalised squares.  A reasonable estimate indicates that
for 6x6 squares it will take about 23500 seconds on my computer.  But I
didn't test it, because that's still longer than I am prepared to wait for
it :)

Kristof