Re: [eigen] compile time statistics

[ Thread Index | Date Index | More Archives ]

Thanks a lot! Very, very, very useful.

I'll have a look at my worst recent commits, 813931 and 815061. As for the 
recent oscillations around 821090, it's just that a bug compiled faster, 
nothing to see here.

> for the cost model bench note that the number of loops depends on the
> size of the matrices, so you cannot  really compare 2 "rows".

ah ok, i'm stupid.

> Moreover 
> the number of loops for the scalar product is divided by two. The
> purpose was to get a similar running time for all cases. That is weird
> is that the scalar product is mush slower than the matrix sum, maybe
> we could try a cwise product to see what happens....

perhaps it's just that muls still cost more than adds...


Attachment: signature.asc
Description: This is a digitally signed message part.

Mail converted by MHonArc 2.6.19+