Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!mips!apple!portal!cup.portal.com!jpser
From: jpser@cup.portal.com (John Paul Serafin)
Newsgroups: comp.sys.handhelds
Subject: Re: HP48sx Simulator for Dos?
Message-ID: <41569@cup.portal.com>
Date: 23 Apr 91 03:02:23 GMT
References: <1991Apr20.012113.27024@qualcomm.com>
  <1991Apr20.165115.27895@bronze.ucs.indiana.edu> <hysky.672222998@polaris>
  <1991Apr21.184105.8597@nntp-server.caltech.edu>
Organization: The Portal System (TM)
Lines: 24

madler@nntp-server.caltech.edu (Mark Adler) writes:
>Doesn't have to be, especially on a modern, fast processor.  I have a
>286/287 emulator running on my 25 MHz 68040, and it's speed is equivalent
>to a 10 MHz 286/287!  Pretty impressive.  (It's called SoftPC from Insignia.)
>Since the 2 MHz, 4-bit processor in the 48 is far slower thatn a 286, I'd
>venture to say that it is well within the realm of possibility to write
>a Saturn emulator that runs *faster* that the real one on, say, a 25 MHz
>386.
If an 8088 is a 16 bit processor and an 80386sx is a 32 bit processor, the
Saturn processor is 64 bits.  The Saturn was designed to crunch numbers
and has several 64 bit registers.  The only thing 4 bits wide is the
bus.   It is a smart bus is addition; in the HP-71 most devices on the
bus kept their own copy of the instruction pointer.  This significantly
reduced bus traffic.   On number crunching benchmarks, a 600kHz HP-71
runs as fast or faster than a 4.7MHz 8088 running interpretive Basic
(the HP-71 runs interpretive Basic).  The 8088 will smoke an HP-71 on
empty loops and integer trivia, but when was the last time an empty loop
did any good?   They certainly cause a lot of harm in software written 
for an 8088 running on a 286, not to mention 386 or 486.
This is not to say that a BCD HP48SX emulator on a 386/387 or 486
wouldn't be very nice to have.   There is a company that markets
Rocky Mountain Basic for MS-DOS but I don't know much about it.
John Serafin
jpser@cup.portal.com