Subject: Re: [boost] : different matrix library?
From: Maurizio Vitale (maurizio.vitale_at_[hidden])
Date: 2009-08-27 11:54:12
>>>>> "DE" == DE <satan66613_at_[hidden]> writes:
DE> on 26.08.2009 at 20:38 joel wrote :
>>> a thought about operations tweaks
>>>
>>> one may want to tile and/or partially unroll operation loops but
>>> since such thing characterize operations rather than objects
>>> themselves it is wrong to tag the objects with such info
>>> consider
>>>
>>> matrix<double, tile<3, 3> > tiled1; //the first one //...
>>> matrix<double, tile<3, 1> > tiled2; //far from the first one
>>> //... matrix<double> result = tiled1*tiled2; //what tiling
>>> would occure
>> In those cases, we compute the smallest common multiple of tiling
>> shape as stated in all loop optimization techniques paper.
DE> i consider it a bad (erroneos) practice because in such a case
DE> neither of programmers intentions would take place but something
DE> totally different instead
DE> '3' is a common number as well as 'power of 2' in such a common
DE> case like
Seems like you're determined to keep this thread alive beyond
reason. May I humbly suggest that at least you do the following before
posting:
- read the text you're answering to
- make yourself familiar with the relevant literature
Had you done so, you would, maybe, discovered that:
- a common multiple is not a common number, although they have
common in common.
- more, a common multiple of a tiling shape is not even a number
(although it can represented by one in R^n).
- last, that people have written about this profusely and
googling for a few keywords would be a good idea.
Other than that, yes, 3 is a rather common number. Not as nice as my
personal favourite, 42, though.
Best regards,
Maurizio