Re: Structured Programming using Forth

On Apr 3, 7:01 pm, John Passaniti <n...@xxxxxxxxxxxxxxxxx> wrote:

In the old version that had only one processer per chip, they were
hoping to get the sort of volume that would let them sell for a dollar
a chip. If they can get the cost down now, if the industry has
progressed to the point they can sell for a dollar or two a chip
today, then the issue isn't how to *most efficiently* map your
computation to multiple processors etc.

This statement doesn't make sense to me. You seem to be making some
equivalence between getting a bunch of sub-buck processors, inventing
the glue and communications scheme between them, working through some
means to boot, debug, and test the whole mess... and the SEAforth chip
which (presumably) already addresses these issues.

No. But first I should point out that I'm not really up on the
competition here. I'm vaguely interested in Forth chips because I'm
interested in Forth. Jeff Fox can make wonderful claims for Forth
chips and I wouldn't know whether he's claiming things that are
routine for competing nonForth chips.

Jeff claimed that the last 1-processor Forth chips were way ahead of
the competition, that they did a whole lot with a little bit of real
estate, that they could produce very large numbers from a single die,
that they were way way ahead of anything else that would be in their
price range once they were produced in large numbers. But they were
never produced in large numbers. iTV didn't go into production and no
other customer was ready to pay for large numbers of a single-source
proprietary chip, and it never happened.

I'm going to suppose that Chuck has kept improving chips so that the
new chip will also be far beyond anything in it price range. A chip
with 25 processors will cost less than 25 times as much as a chip with
1 processor because things like pins and packaging are part of the
cost, and also things are smaller now, so I will optimisticly suppose
that produced in large numbers they could bring it down to a couple
dollars a shot.

Of course I'm being optimistic.

The problem is to get your work done. If you see a way to use the
extra 24 processors to improve your result, then go for it. If you see
a way to use a second processor, that's good too. You're under no
obligation whatsoever to optimise your use of 25 processors. Your
first obligation is to get a working product out the door, and if it
turns out to be easiest to do that using 4% of the computing power on
the chip, that's fine. Expand into using more processors as you see
how.

This is very naive.

When I as an engineer am asked to choose a processor for a product, I am
expected to choose a processor appropriate for the task. If the task
only takes 4% of the processor, then it is not appropriate.

No, it's appropriate if it will do the job you need done and it's the
cheapest alternative factoring in all your expenses.

Back when computing power was expensive, when an hour of computing
time cost ten times as much as n hour of a programmer's time, you
could spend 5 hours reducing a run-once program's time from 2 hours to
1 hour and your cost was justified. Those days are gone. Now you can
afford to wast 96% of your processor time provided the job gets done.

Similarly, if your boss is deciding whether to use you on a project,
he doesn't think "Is this going to use all of John's abilities? I'd
hate to have John doing work that doesn't challenge him completely.".
But he might think "Is this project so easy that I can trust it to the
guy who gets half John's rate?". He doesn't care that you know a
hundred languages and will use at most 5 on his project. He cares that
you can do a superb job in half the hours.

So you *are* under an obligation to use as many of the processors as you
possibly can in the most efficient way you can. Not using 96% of a
processor's bandwidth means there is very likely some other solution out
there that is less expensive. As an engineer, that is one of my primary
roles: to find ways to do more with less.

If doing more with less is vital, you could go to Intellasys and ask
for a custom chip with just 1 processor, or just 5 processors, the
minimum you need. Maybe that will fit your needs better than the 25-
processor chip. I'd expect you'd do better to find a cheaper
alternative from some competitor. What you want is something cheaper
that will do the job -- you don't care about using the weakest
processor that you can get by with unless it's cheaper, right?

I once had a chance to look at a robotics application, a machine that
carried blank silicon wafers from one site to another in a clean room.
The machine was of course very very expensive. And all its various
functions were managed by two 6509s. They wanted it to do fast
accelerations and decelerations without vibrating the wafers, and they
were using hours of engineer-time to do a square root faster on one
6509. One of the other guys on the tour was trying to figure out how
much meny they could possibly be saving by not using a more powerful
chip. For ten dollars more, how many bottlenecks would you widen? (But
I don't know their criteria, there might have been an important reason
to use only proven chips inside the clean-room.) A low-volume high-
cost application, and they were using a very cheap part that was
barely capable of getting the job done.

Everyone is getting excited over the dribbles of information from
Intellasys. I'll get excited when I see the tools, models, and training
that they offer to actually make their hardware live up to the promise.

Sure. It's particularly interesting because Chuck is doing it, and
he's done great things before. Pioneering things. "You can tell the
real pioneers by the arrows in their backs." Chuck doesn't have a
great track record for projects that turn commercially successful
right away.

You don't sell chips because you're a pioneer. You sell chips because
you have something more compelling than the other options. You don't
sell chips because your name is Mr. Moore and you have 24 processors on
a chip. You sell chips because you've not only thought out the hardware
but how real-world programmers trying to solve real-world problems will
actually use your chips.

Yes, exactly. *I'm* interested because it might be pioneering work by
Mr. Moore. You're potentially interested because it might be something
you can actually use. If I hear about some neat trick I'll be
interested in it because it's a neat trick. You'll be wondering
whether it's useful to you in the near future.

Intellasys needs to pay attention to the needs of people like you. All
people like me are good for to them is to perhaps get the attention of
people like you -- the more open-minded ones, who won't be put off by
the source.

.

Relevant Pages

  • Re: Richard Stallman is responsible for the shrinking economy
    ... product than when you try to sell it on its own. ... get the perception that Sw is valueless. ... With proprietary tools, you can't even fix the problem. ... vendor, more choice means more chip sales, so they encourage multiple ...
    (comp.arch.embedded)
  • Re: Maxim parts and lead times.
    ... > When I request samples, I usually get them in a week or so. ... > common MAX232 chip takes two months to purchase! ... them to agree to split the reel and sell me 200, but even that was too much ...
    (comp.arch.embedded)
  • Re: Elvira wont talk but the other guy does
    ... My board does have the U6 chip. ... it is the board out of my Whirlwind (the Whirlwind sound board with all ... Plus it had legs from two different machines, no translight glass for ... been able to sell it on EBAY for that price knowing what someone would ...
    (rec.games.pinball)
  • Re: Meade Deep Sky Imager
    ... I'm willing to be if Canon made their 6.3 Mpixel chip/electronics available ... a peltier cooled USB2 camera based on that chip would ... sell like crazy. ...
    (sci.astro.amateur)
  • Re: AMDs 45nm technology compared against Intels
    ... chips that used to cost 50 dollars now cost 50 cents. ... If a chip uses 100 mm2, there are possibly about 700 chips on the wafer. ... When manufacturing starts on that 100 mm chip, yield sucks and you only get a few good ones per wafer so they cost hundreds of dollars each to make. ... You must admit Intel has done some market manipulation on occasion. ...
    (comp.sys.ibm.pc.hardware.chips)