|
| Authored by: argee on Friday, May 10 2013 @ 10:23 PM EDT |
Earlier cpu chips were very limited in the operations they
could perform, therefore entire software libraries were
implemented to perform things like subtraction, multiplication, division,
decimal calculations, etc.
With the appeals court decision, those libraries would NOT
be eligible for patent.
However, if Intel, AMD, Zilog, etc. invented a new CPU that
had the hardware within to, say, multiply, then THAT cpu
would be patent eligible.
I remember in the 8080 days, I would have simply KILLED
for being able to load register A with a number,
register B with a number, and then issue the assembler
command multiply A x B and put the answer in register
(take your pick).
My assembly coding ended with the Z-80, so I do not know
if this was ever done; I am sure it has. I remember the
joys of reading a sector from a floppy. Not trivial with
an 8080.
---
--
argee[ Reply to This | Parent | # ]
|
| |
| Authored by: Anonymous on Sunday, May 12 2013 @ 07:19 PM EDT |
Each instruction in a CPU's instruction set can and is represented as a
mathematical formula. Since a program is a series of these instructions, it
stands to reason that a program made of a series of these instructions could be
represented by the mathematic formulae that defines the function of the
instructions of the compiled program. One could even write a program with the
guts of a disassembler that decodes the object code, takes the mathematical
definition of the instruction set, and prints the mathematic formulae for any
program it opened. The printout would be large, but it would be 100% maths.
The field of mathematics is large, but the component executed by a CPU is only a
small subset. That does not mean it doesn't do maths, it just means that it only
does a small part.[ Reply to This | Parent | # ]
|
|
|
|
