At 02:11 PM 10/9/2001 -0400, Michael Maraist wrote:
>Oh by the way.
>
>500M ops in 15 seconds is 1.02 Billion parrot ops / second. This isn't
>too shabby on a mere 466MHZ machine.
The performance is impressive regardless, but isn't that 33M parrot ops/sec
rather than 1.02G ops/sec?
>Pure-mathematical real-parrot-ops should
>get pretty close to this, but most parrot-ops will have evil conditionals
>that thwart such performance.
Stamping out evil conditionals is one of the reasons for the variable
vtables. We can't get rid of *all* the conditionals (it'd be a dull program
that didn't make some decision... :) but if we're going to blow pipeline we
might as well do it once and real big.
Dan
--------------------------------------"it's like this"-------------------
Dan Sugalski even samurai
[EMAIL PROTECTED] have teddy bears and even
teddy bears get drunk
