But if a designer makes a CPU that is neither mnemonics compatible nor
binary compatible, but that uses the same ideas (e.g. all opcodes are
conditional, barrel shifter into the data processing instructions, etc..)
can this infringe patents too?

I'm afraid I already guess the answer.. :/


At 10.23 2011.10.17, you wrote:
>It is my understanding that the instruction set was patented making it=20
>an infringement to create a binary compatible core.
>
>There are some neat features in the ARM that came about from some=20
>great thinking by the BBC micro team, many of whom are still around=20
>Cambridge but not at ARM anymore.=20
>
>Conditional execution is a neat feature but it does not seem to be=20
>used as much as it could be.
>
>On 2011-10-16, at 9:14 AM, Electron wrote:
>
>>=20
>> Hi all,
>>=20
>> now that the original ARM patents are (supposedly!) expired, anyone=20
>can design
>> a CPU (I do not mean copy an existing ARM core, but design one from scra=
tch)
>> and not incur in patent infringment?
>>=20
>> And, when the original ARM patents were still in place, was it legal=20
>to design
>> and sell a CPU binary-compatible with ARM (not licensing anything of cou=
rse)?
>>=20
>> This of course depends if the opcodes were part of their patent, but ass=
uming
>> they were..
>>=20
>> Just (very) curious. Of all the CPU's I programmed for (many), I=20
>think the ARM
>> has the most beautiful instruction set. Expecially the "all instructions=
 are
>> conditional", although not exclusive of ARM, is just so nice on it.
>>=20
>> Thanks,
>> Mario
>>=20
>> --=20
>> http://www.piclist.com PIC/SX FAQ & list archive
>> View/change your membership options at
>> http://mailman.mit.edu/mailman/listinfo/piclist
>
>
>--=20
>http://www.piclist.com PIC/SX FAQ & list archive
>View/change your membership options at
>http://mailman.mit.edu/mailman/listinfo/piclist

--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.