Re: HP 41 CPU speed Message #3 Posted by Ex-PPC member on 6 Feb 2002, 4:32 a.m., in response to message #2 by John Ioannidis
John wrote:
"The buses ran at about 355KHz, but each machine cycle
was 56 clock pulses long; the machine would execute
about 6300 machine instructions per second. [ ... ]
The calculator had just 12K of 10bit words worth of ROM
Amazing, no?"
Yes. However, not being a hardware man, I find it
curious that the HP-71B had just twice as fast a clock
(about 600 Khz, if I'm not wrong), yet it was much more
than twice as fast as the 41C, up to 14 times or more.
I guess the reason is that each 71B's machine cycle
was much shorter than 56 clock pulses long ? Or was it
because of an improved architecture, or a combination
of other facts ?
Another fact I find curious is that the 41C could
execute some 6300 machine instructions per second. But
it could compute quite complicated trascendental
functions such as e^x or tan(x) in well under a second,
so I guess this means it had to execute significantly
less than 6300 machine operations to compute them ?
Hats off tehn to the algorithm designers and coders !
|