From: Ross Archer <archer_at_topnow.com>
Date: Tue Jan 30 00:42:11 2001

"Claude.W" wrote:
> There was also a beige all plastic model that came later. Cheaper to make.
> These seem a bit more rare then the silver/black ones....here anyways...
> A fast 16 bit CPU but a weird way of accessing memory made the thing really
> slow IIRC

Correct me if I'm wrong, but I've heard the 99/4A suffered something of
unfair rap for being slow due to the fact that its TI BASIC was
inefficient and sloppy, and that little or no blame for its slowness
on the actual CPU or hardware. (Windoze users may spot a parallel here.)

The TMS 9900 CPU actually used a "window" in RAM as its register space,
an internal pointer register to locate its base address. This was
a great idea at the time, because CPU cycles were long enough back then
it made no difference whether you stored temporary values in internal
or external RAM, as either could get you your data in the 1 cycle time
(And interrupt latency can be really short if you can make a "fresh"
register bank
with one register load!)

Most CPUs like the 8080 had barely enough registers (and special-purpose
ones at that),
so it was an advantage to use the wide-open RAM space for storage
instead of doing
a lot of register-shuttling XCHG-type stuff working with a tiny register
It's one of the reasons that
6502 code density was so high, despite having obvious holes in the
instruction set,
because zero-page was effectively 256 byte registers and/or 128 16-bit
indexing with them with two byte instructions. And why Z80 added
indexing and
extra register banks -- they were sorely needed.
(/me - 8080 and 6502 programmer from way back. Talk to me about XCHG,
XTHL, and all kinds of neat tricks. It was fun but despite, not because
of the CPU

Of course, modern RISC goes back to the load/store, many on-chip
register model, but
isn't it interesting that Berkley's SPARC uses register windows much
like the TI

-- Ross
Received on Tue Jan 30 2001 - 00:42:11 GMT

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:33:48 BST