Re: matrix operations and temporary objects?

[Leo Razoumov <slonik.az@...>]

On 8/4/09, steve donovan <steve.j.donovan@gmail.com> wrote:
> On Tue, Aug 4, 2009 at 9:28 AM, Asko Kauppi<askok@dnainternet.net> wrote:
>  > We're currently using Lua for matrix calculations with matrix sizes around 50x60 to 500x600. Operations are SSE optimized, but we do not do any operation merging or GC avoidance.
>
>
> So the question is, is worrying about this premature optimization? Do
>  we have some numbers?
>
>  steve d.

In numerics this problem is common knowledge that was beaten to death
20 years ago and providing any performance numbers at this point would
not add much value. If you need to do few matrix ops now and then you
typically do not worry about performance at all. But if you run it in
an inner loop billion times per simulation you do care about billion
extra mallocs and GC cycles.

--Leo--