|
Message
From: Bill Cox<bill@v...>
Date: Thu May 13 11:24:45 CEST 2004
Subject: [oc] One issue about free hardware
Hi, Rudi.> > I wonder what a good term would be for the things that are currently described
> > as 'IP' in the hardware field. Ie a common phrase is something like 'the memory
> > controller is someone else's IP'. Maybe proprietary core vs free core?
>
> Thats actually a very good point. Perhaps this is a good
> time to create a new term.
>
> As I understand from Richards, post the words "intellectual"
> and "Property" are bad choices. So perhaps we should try to
> come up with a catchy name that can replace "IP".
>
> How about:
> 1) "Functional Block" or "FB" for short
> 2) one variant would be "Free Functional Block" or FFB
I remember how I first reacted to the term "IP" as it's now used in the
hardware cores field. I thought it was pretty dumb. I read it in
EETimes. Guess who gets to define these terms? Not you and me, but the
authors, and evil marketing guys at large companies.
Frankly, unless we're going to have a huge marketing campaign (and the
press goes along), it's too late to redefine the term IP. Other names I
hate: 9/11. It's a date, not an attack. It also sounds like a
Porsche. AI was a real winner. Imagine trying to duplicate human
intelligence with a PDP-11. How about "custom" ASIC? Isn't that a
contradiction? How about how we say a million dollars? I write $1M.
The financial guys write $1MM. What's MM? Candy! How about FPGA vs.
CPLD? Someone needs to beat up Altera and convince them that they
actually make FPGAs.
Here's a story about what happens when we keep changing the name of a
thing...
For years I tried to sell structured ASIC technology to silicon
vendors. We had to come up with our own term, since there was none in
common use. We threw our hat in with Chip Express and called them
"Module based arrays". Lightspeed used a different term, just so they
could pretend they invented the idea, and then of course LSI used yet
another, and they still refuse to call their technology Structured
ASICs. The silicon vendors though we were nuts for trying to sell them
a thing they couldn't even describe with a simple phrase.
My understanding is that the guys at EETimes got tired of our terms, and
started calling them structured ASICs, which I think is a dumb name --
what does "structured" mean? Does it imply that traditional ASICs have
no structure? The term is only a little more than one year old, but now
everyone knows what it means. Now when I talk to potential clients,
they know what I'm trying to sell. Getting a name for a thing is huge.
The technology is at least 15 years old, but it took 14 years to come up
with a name!
While I'm bashing names I don't like, how about GNU? It's cool for hip
college kids, but the term doesn't stick in the brain easily. Linux is
much easier to say, and easier to learn (at least when pronounced like
"lin-ux", rather than "line-ux"). Why do you think everyone says "I'm
using Linux" instead of "I'm using GNU"? And what in the heck is that
animal on the GNU web page? I bought a Linux Penguin for my daughter.
She even calls it "Tux". I'm not sure I'd buy a toy animal I can't even
describe. I know... it's got a name, I just don't know it!
Unfortunately, all these terms I hate are what they are for (mostly
evil) marketing reasons. Trying to fight them is IMO a lost cause.
One term I like is RMS. It's not only a way to measure signals, it's a
guy (Richard Stallman). Wow. He's got his own word! Now that's a
marketing achievement. Also, it makes me feel cool, since I'm one of
the in-crowd that knows what it means (ok, I don't know what M stands
for). Even Bill Gates doesn't have his own word.
Bill
|
 |
Browse
