Re: new HP CPU???



#4

With CPU speeds generally so far beyond single application requirements, I wonder what would be the justification for a truly new CPU?

If HP has made a decision to switch to a new technology, it's already surrendered at least some of the value of its investment in the current code base. Why create new technology when there's so much already implemented, available and supported in the form of fast, low-power versions of the Pentium?

What's the justification? Is it because they're implementing single-chip solutions and (presumably) need their own design anyway? Are profit margins so slim that they justify implementation rather than licensing?

I'm not (strictly) in the business, so maybe I just don't get it. Anyone care to comment?


#5

The obvious upwards path to more power would be to create a faster Saturn chip.
However, I faintly recall that it was mentioned someplace that for the HP48G, HP did push the Saturn to its limits, that is 4 Mhz.
Can anyone shed some light on this issue ?
Is it related to the nature of the chip, which might be an hybrid or something ?


#6

Just guessing, but some ideas and sub-ideas:

1) Only justification for a new processor is to have the same or very similar instruction set, so not have to rewrite the assembly code for the calculators.

2) If the firmware is coded in other than assembly (i.e.:RPL), then the software could be migrated to any processor in which a kernel capable of supporting RPL could be developed.

3) As new processors are much faster (my pocket PC has a 206 MHz StrongARM RISC Processor, just to talk about a portable product), there could be a software emulator that emulates the original instruction set running on such new chips, and attain acceptable performance.

4) The HP LX 100/200 are good examples of an excellent HP calculator application ported to a 8086 processor. So code is available (for HP, I mean, because of their intellectual property)

5) Good emulators could run in a Java Virtual Machine, attaining high portability. There are examples here at the Museum. So many processors could "host" a calculator.

6) It is very possible that the cost of manufacturing today a chip designed 15 years ago (approx.) may be higher than the cost of manufacture for a current design, compatible with today's batteries, displays, voltage levels, manufacturing techniques, current test equipment, etc...


Possibly Related Threads...
Thread Author Replies Views Last Post
  HP-71B CPU frequency? Christoph Giesselink 18 3,534 03-02-2013, 09:35 AM
Last Post: Gerson W. Barbosa
  OT: Mechanical CPU clock Ethan Conner 1 882 04-19-2012, 01:23 AM
Last Post: Mike Morrow
  HP 15C LE programming, does CPU run fast? designnut 1 838 02-12-2012, 05:13 PM
Last Post: Jeff O.
  Emulator CPU usage Martin Pinckney 6 1,619 10-13-2010, 01:43 AM
Last Post: HrastProgrammer
  Capricorn CPU microcode Olivier De Smet 6 1,589 03-10-2010, 03:50 AM
Last Post: Olivier De Smet
  HP nut CPU and illegal instruction sequences Håkan Thörngren 9 1,889 01-05-2010, 12:50 PM
Last Post: Eric Smith
  Lab-Volt (348?) cpu trainer peter a walker (papwalker) 0 593 09-02-2009, 08:48 AM
Last Post: peter a walker (papwalker)
  P41CX, Palm TX, and CPU Speed setting Les Wright 8 1,601 04-27-2006, 07:29 AM
Last Post: Thomas Okken
  HP 35 CPU Tom Bagdas 6 1,367 09-02-2005, 07:45 PM
Last Post: don wallace
  HP 12C CPU Question Jeff 0 521 12-13-2004, 10:47 PM
Last Post: Jeff

Forum Jump: