Mail Archives: djgpp/1999/02/02/19:55:46
At 03:09 PM 2/2/99 -0500, you wrote:
>OK, I was close. But, the 387 *can* be used as a 64-bit integer math
>unit, as it has 64-bit int load/store, and if all the numbers and
>results are 64-bit integers (i.e. be careful with division,
>multiplication, arctangents, etc), you can treat it like a 64-bit
>integer unit.
>
>It's not as fast as simulating it in the CPU though.
Really? "simulating" it? Is there no 64 bit (integer) multiply in the Intel
alu?
What is the best that an intel type chip can do for 64 bit int operations:
Intel Cyrix AMD
-- Pentium -- 5x86? K5?
-- PPro -- PPro MMX
-- PII -- PII MMX --Celeron -- 6x86 K6
-- PIII
Also, is the 64 bit long long arithmetic provided by GCC invariably as fast
as 64 bit int arithmetic can ever be on a given platform? (Amortized,
assuming the bestpossible compile optimization -- e.g. PGCC on a Pent class
chip, with the best arch setting for the particular chip, e.g. -mamdK6.)
--
.*. "Clouds are not spheres, mountains are not cones, coastlines are not
-() < circles, and bark is not smooth, nor does lightning travel in a
`*' straight line." -------------------------------------------------
-- B. Mandelbrot |http://surf.to/pgd.net
_____________________ ____|________ Paul Derbyshire pderbysh AT usa DOT net
Programmer & Humanist|ICQ: 10423848|
- Raw text -