I've done some timings on SBCL v. GCL bignum routines; GCL is faster --
and not just because SBCL fixnums are only 30 bits v GCL's 32 bits.
I don't know who the maintainers of SBCL are, but their bignums could
be made faster.
Also, GCL's type info is terrible! GCL doesn't even realize that
BIT's are only 0's and 1's! E.g., (typep 3 'bit) => t !! (GCL 2.6.1).
That may account for some of GCL's slowness == inability to know how
bit things are.
BTW, which Common Lisps support ~ 64-bit fixnums & address spaces > 4 GBytes ?
At 11:55 AM 3/24/2013, Rupert Swarbrick wrote:
>Raymond Toy <toy.raymond at gmail.com> writes:
>> Have you considered how fast or slow this is on older machines? It
>> seems that most developers have reasonably fast machines so we might
>> not notice. But I think some (many? few?) users have much slower
>> machines. Or at least some had machines without sse support, which is
>> a pretty old machine.
>
>That's why I wrote up the timings carefully. I intend to spend some time
>over the next few days seeing if I can speed up the GCL and SBCL
>versions a bit more. Unfortunately, it doesn't look like the ECL version
>can run faster than double the speed it currently does (which is why
>there's the note about timing reading line by line in the commit note).