Re: never enough
Dear MISC readers:
"Richard S. Westmoreland" wrote:
> > I still consider going to FPGA a step backwards. Chuck decided to go
> > from PGA to custom silicon in 1989 to get an order of magnitude or
> > two speedup and two orders of magnitude improvement in production
> > cost. This is a 1000/1 or 10,000/1 ratio in price performance.
> > It is a very big hit to take. The only justification that I can
> > see would be not having enough money for custom which is more
> > expensive. But the 1,000/1 or 10,000/1 ratio is hard to ignore.
>
> Yea... doesn't that evade the purpose of the minimal instruction set
> architecture?
FPGA do carry fat in comparison to custom VLSI, but I don't think
that the use of FPGA is contrary to the concept. Chuck began with
PGA 16 years ago and eventually moved to VLSI. I ended up in a
position to jump in on VLSI rather than FPGA and I understand
why Chuck wanted to go that direction. But I also understand
that FPGA is more practical place to start and that it does offer
faster turn around and lower development cost. But the concepts
can still be there. Custom I/O circuitry, tight integration,
custom MISC architecture and instruction set etc.
Chuck is a perfectionist and wants to push a given process as far
as he can, but most other people just spend more and move to a
faster process where efficiency will not be as important. Just
because Chuck has used older technology available at prototype
fab companies doesn't mean that that is part of the MISC concept.
It was just the best thing available to us at the time.
Other things will be available to us and to other people at other
times and the concepts of what we are doing can still be applied.
I can understand going with FPGA when starting from scratch but
because I am still working on a custom VLSI chip I don't see
the same advantages as other people.
I have more than 40 Forth chips listed on the
http://www.ultratechnology.com/chips.htm page now and many of
them are FPGA designs. As I say, I don't consider FPGA stuff
to be in any way contrary to MISC philisophy. I just think
that it is a step back for companies alreading working in
custom VLSI. Though not FPGA the 4 bit Forth microcontrollers
with "many times programmable" memory listed on the page
look really nice to me for very tiny embedded stuff. Talk
about Forth stamps!
But I do also acknowledge that FPGA is also a step in a different
direction. What is back to us is forward if you are heading
for a different target. There are prefectly valid targets for
designs that take advantage of FPGA technology.
> Videos on DVD would be nice, but I wouldn't sell them for any less than $300
> a piece.
For that I could buy a player. ;-) No really, I don't think that there
is
as big an audience for DVD as CD. Many people have CD players now on
their computers, some have DVD but the number is smaller. The main
point is that CD media is cheaper than tape media when you factor
in all the costs of wear and tear on equipment, time, packaging and
shipping etc.
When I was at iTV we looked at CD-ROM hardware. It is remarkably
small and simple. I can understand why a CD-ROM drive can be cheaper
than a floppy drive (lightweight components and read-only hardware)
We thought it would be nice to make something the size of a moneyclip
that is a computer that reads or writes CD and DVD. There are nice
business card sized CD-R with 50MB. A business card sized computer
that holds busines card sized CD would be cute.
But if a DVD video was going to be $300 I would want to see some
pretty impressive special effects or a director's cut with some
big name stars. We have Chuck, Ting, John Rible, Skip Carter,
Wil Baden, myself, and others, but the tapes are like home
movies, not high budget Hollywood style. :-)
I seem to be able to make originals and 100 video tapes in the life
of a camcorder, so factor in say $8 for video equipment. A master
tape for the initial recording is $18 and if five tapes are made
that is $3 per tape. high quality VHS blanks are $10. $2 for
a shipping envelope and $3 shipping is now $26. So if I charge
$35 shipping and handling I am getting $9. It takes about three
hours to make a tape, and package it, four if you count a walk to
the post office. It comes to $2 or $3 an hour at those rates.
I lose money on tapes in PAL format due to the extra costs and
would lose money on tapes if I charged much less.
With a CD-R I could provide cheaper media and duplication might
be faster and would not add to wear and tear on the camcorder
and VCR, just on the computer and CD-R (which I don't have).
Packaging and shipping would be less, and I could thus offer
videos on CD-ROM more easily than tape if and when I get
a CD-R and some blanks.
I figure I will offer videos on tape, CD-ROM in mpg format,
and small versions in .rm format on the web. mpeg also
makes it easier to make transcriptions since it does not
cause the wear and tear on a camcorder or vcr that starting
and stopping the tape and replaying sections over several
times causes. But it is easy to fill up a hard disk with
a few mpg video files so a CD-R, CD-RW or DVD-R seems
like a good match to the size of mpg files.
But as I said, whatever is never enough. Whatever we have we will
want more. Faster chips, higher resolution video, faster interconnect
bandwidth, more chips, smaller geometries, faster clocks, better
software etc. Things tend to move forward but never as fast as
we would like.
Jeff Fox
