7 Mar
2011
7 Mar
'11
4 p.m.
Chuck Guzis wrote:
There were math co-processor boards (AM9511, AM9512,
for example) for
S100 systems. ISTR that there was even a TRW bipolar (16x16? Huge
chip that ran hot as a pistol) for the S100 bus.
I suppose if it appears transparent to the user (machine executes instruction
xx, doesn't know what to do with it, and throws it at the copro) then it's a
valid solution. I think I'm more interested in systems that came with support
as standard rather than as an optional cost extra, though.
The Moto 6809 had an 8x8 multiplier.
Yes, that seems to be the 'famous' one that gets mentioned everywhere. It
seems it was of the shift-add variety. Anyone recall if it would work with
signed integers? (I'm just trying to work out how the math works for signed
multiplies at the moment)
The TI TMS9900 had both multiply and divide
instructions.
Ahh, that's a useful one to know.
All mainframes of the time that I'm aware of had
floating-point
multiply. Many had floating-point divide (or a mechanism for getting
there). Many lacked integer multiply and divide.
Any particular reason that they'd support FP ops but not do the same for
integers? Does that imply that it was typically done by an (optional?) added
FP 'math' unit rather than being a core part of the CPU?
cheers
Jules