Re: A86: Serious Multiplication


[Prev][Next][Index][Thread]

Re: A86: Serious Multiplication




In a message dated 8/5/98 6:02:39 PM Central Daylight Time,
assets@eden.rutgers.edu writes:

> Here is a fast multiplication routine.  In its worst case, it runs
>  through the loop 8 times.  Multiply any h by any l and the result is
>  returned in de; this might be fast enough to drive ray-tracing :-)


Great!  I wonder if this is faster than _FPMULT.
Now you have to do it again for division.

Also, ray-tracing is so cpu intensive, that it's extrememly slow on even
today's fastest systems.  I think you meant ray-casting, didn't you?

-Dave