Authored by: Anonymous on Friday, February 15 2013 @ 04:59 PM EST |
Your telegraph example is also well taken, thanks.
It will be on my mind...
Here's another distinction this discussion brings to mind
which may be more on-point with respect to patent materials.
I start with three FPGAs. On the first, I have programmed
a mechanism which acts as a shift register (to keep with
a prior example). On this hardware, it can shift bits at
500MHz. As I reach for the second chip, my co-worker says
"No, wait, try this one, it's faster." so I grab his and
program it with the same synthesis. Indeed, it now runs
at 600MHz, and I am very pleased. My other co-worker,
who tends to think a bit more deeply than me, has been
looking at my design and says "Oh wait, look here, let's
cross-connect the gates in this clever fashion, so that
the shifted value can be read out without actually
shifting the bits." We spend an hour or two on the code
and reach for another FPGA, this time the slow version.
Programmed with this new image, we find that this shift
register now runs at 600MHz >on the slower hardware<!
The sticky labels on the top of the chips that say
"1", "2", and "3" obscure the printed-on numbers
that
would tell us which is the faster hardware. Scratching
our heads, we try to remember, was it "2" or "3" that
had the fast clock-to-output numbers? The chips look
exactly the same, the pins go up and down in the same
order at the same rate, and both stop working when we
try to run them at 601MHz. For practical purposes, they
appear to be identical.
The question is, which "chip" is patent-qualified
intellectual material? One, both, neither? I'll
certainly appreciate your comments.
-jrl
[ Reply to This | Parent | # ]
|
|
|
|