Intellasys Site Has Changed!
Jason Damisch
It says that thumb drives with the Forth chip in them are now
available. :^)
The login area is gone.
The site has a new look.
There are no jobs available at the moment.
There is a message board to post onto.
Jason
John Passaniti
> http://www.intellasys.net/
>
> It says that thumb drives with the Forth chip in them are now
> available. :^)
Kinda. It says "available on request to qualified parties." And the
datasheet on the SEAforth chip is no longer available for download; you
now have to write with a "brief description of your application."
Looks like they may not have large quantities of the thumb drives
available. As for the datasheet, someone there probably wants to build
a sales contact database.
> There is a message board to post onto.
With some code snippets.
Elizabeth D Rather
Intellasys is trying to focus their sales and support efforts on
potentially significant users of the technology (folks who will buy
large quantities of parts). That's a common business model for
companies introducing new and innovative technologies. Harris Semi
followed a similar plan in the 80's with the RTX parts.
Cheers,
Elizabeth
--
==================================================
Elizabeth D. Rather (US & Canada) 800-55-FORTH
FORTH Inc. +1 310-491-3356
5155 W. Rosecrans Ave. #1018 Fax: +1 310-978-9454
Hawthorne, CA 90250
http://www.forth.com
"Forth-based products and Services for real-time
applications since 1973."
==================================================
John Passaniti
> Intellasys is trying to focus their sales and support efforts on
> potentially significant users of the technology (folks who will buy
> large quantities of parts). That's a common business model for
> companies introducing new and innovative technologies. Harris Semi
> followed a similar plan in the 80's with the RTX parts.
And we all know how well that worked out.
Sales: Many semiconductor manufacturers are now using distributors
like Digi-Key and Mouser for samples and evaluation boards, allowing
their representatives to focus on volume sales and support services.
I have to believe that it would be pretty trivial for Intellasys to
set up a relationship with one of those distributors for samples and
evaluation boards (like the thumb drives, like the home theater demo
board they talked about in the past). Their sales staff wouldn't have
to deal with such trivia, while at the same time, these chips would
start to get into the hands of engineers who wanted to evaluate them
and hobbyists who wanted to learn about them.
Support: I don't get this either. They are going to have to produce
documentation anyway (datasheets, application notes, etc.) for their
bigger customers, so why not put those same documents online? What
support for engineers evaluating the chips and hobbyists trying them
out is needed other than basic technical data, and maybe on online
forum where users of the chips supported themselves?
I'm not expecting anyone from Intellasys to hold my hand, and I doubt
most other engineers would either. Give me a datasheet (which used to
be online), some basic tools (which also used to be online), and give
me an evaluation board (like the thumb drives), and I can support
myself. And if I run into a roadblock, my first avenue for support
wouldn't be to go to Intellasys and waste their time, but to check out
online forums (like their own), mailing lists, and so on.
I have to wonder about a strategy of only going for bigger clients.
Seems to me to put all your eggs in one basket, and if one of those
clients decides to switch processors, you're screwed and scrambling to
find someone else. Seems a far better strategy would be to get these
chips in the hands of as many engineers as possible, get them familiar
with the technology, get them thinking about the kinds of applications
they would enable, and then wait for the orders to come in. Get a few
big clients, get a few smaller ones. Get the chips out there, and
engineers thinking about them when they make their processor
selection.
Jerry Avins
...
> I have to wonder about a strategy of only going for bigger clients.
> Seems to me to put all your eggs in one basket, and if one of those
> clients decides to switch processors, you're screwed and scrambling to
> find someone else. ...
It's the marketing version of the Giuliani strategy. I fear a similar
outcome.
Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Elizabeth D Rather
> On Jan 31, 1:28 pm, Elizabeth D Rather <erat...@forth.com> wrote:
>> Intellasys is trying to focus their sales and support efforts on
>> potentially significant users of the technology (folks who will buy
>> large quantities of parts). That's a common business model for
>> companies introducing new and innovative technologies. Harris Semi
>> followed a similar plan in the 80's with the RTX parts.
>
> And we all know how well that worked out.
Actually, it worked extremely well. There were a number of significant
design wins, and orders were exceeding their targets. That entire
division (not just the RTX group) was shut down because an ill-advised
merger elsewhere in Harris caused a major cash crunch.
> Sales: Many semiconductor manufacturers are now using distributors
> like Digi-Key and Mouser for samples and evaluation boards, allowing
> their representatives to focus on volume sales and support services.
> I have to believe that it would be pretty trivial for Intellasys to
> set up a relationship with one of those distributors for samples and
> evaluation boards (like the thumb drives, like the home theater demo
> board they talked about in the past). Their sales staff wouldn't have
> to deal with such trivia, while at the same time, these chips would
> start to get into the hands of engineers who wanted to evaluate them
> and hobbyists who wanted to learn about them.
I think the problem here is that the technology is so dramatically
different that it will take a while (and a lot of orders in the
pipeline) before the distributors can handle it effectively. Most of
the stuff they carry are mainstream parts, or new variants thereof.
> Support: I don't get this either. They are going to have to produce
> documentation anyway (datasheets, application notes, etc.) for their
> bigger customers, so why not put those same documents online? What
> support for engineers evaluating the chips and hobbyists trying them
> out is needed other than basic technical data, and maybe on online
> forum where users of the chips supported themselves?
>
> I'm not expecting anyone from Intellasys to hold my hand, and I doubt
> most other engineers would either. Give me a datasheet (which used to
> be online), some basic tools (which also used to be online), and give
> me an evaluation board (like the thumb drives), and I can support
> myself. And if I run into a roadblock, my first avenue for support
> wouldn't be to go to Intellasys and waste their time, but to check out
> online forums (like their own), mailing lists, and so on.
>
> I have to wonder about a strategy of only going for bigger clients.
> Seems to me to put all your eggs in one basket, and if one of those
> clients decides to switch processors, you're screwed and scrambling to
> find someone else. Seems a far better strategy would be to get these
> chips in the hands of as many engineers as possible, get them familiar
> with the technology, get them thinking about the kinds of applications
> they would enable, and then wait for the orders to come in. Get a few
> big clients, get a few smaller ones. Get the chips out there, and
> engineers thinking about them when they make their processor
> selection.
I generally agree with you here.
Bernd Paysan
> I have to wonder about a strategy of only going for bigger clients.
> Seems to me to put all your eggs in one basket, and if one of those
> clients decides to switch processors, you're screwed and scrambling to
> find someone else. Seems a far better strategy would be to get these
> chips in the hands of as many engineers as possible, get them familiar
> with the technology, get them thinking about the kinds of applications
> they would enable, and then wait for the orders to come in. Get a few
> big clients, get a few smaller ones. Get the chips out there, and
> engineers thinking about them when they make their processor
> selection.
Indeed. A note from my company: When we got a new CEO, he said "we should
concentrate on the larger customers" and stated a 80/20 rule observed from
the current pattern (80% of the volume goes to 20% of the customers). But
he also said that the growth is coming from the smaller customers, and
certainly by dropping the other 80% of the customers, volume would go down
by 20%, and won't grow then.
So what he did is to move the 80% of the smaller customers to distributors
who can do a better job at supporting those (and therefore produce more
growth). He had a learning curve to do that, but he got there, finally.
I think Intellasys has some problems with being open because their mother is
a patent troll. Intellasys on the other hand is an innovative technology
company. There are for sure very different views about how technology works
and how you relate to customers between these two parts (patent troll: the
customer relationship is like highwayman to victim. Concentrate on the big
victims, and keep all your operations as secret as possible ;-).
BTW: Harris did *not* ignore the community with the RTX. The RTX 2000 was
launched in Europe together with the German Forth Gesellschaft, and the
evaluation boards were available through the Forth Gesellschaft (or rather
through members of it, who sold boards like that anyway).
--
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://www.jwdt.com/~paysan/
SEAforth Marketing
Please allow me to jump in and confirm a couple of things said, defuse
some rumors and add some facts. First, my name is Jurgen Krehnke and I
am responsible for Marketing and Sales at IntellaSys.
#1. The Website: We had to revamp the website because it wasn't
structured well before and had over time become unmanageable. We did
use the opportunity to change the look and feel and add the community
forum. This forum is all new and needs to be filled with information,
code examples and relevant topics on our hard- and software. So yes,
it is not the world almanac yet, but we are only starting out! So
please join us, participate and help us make it more useful. On a
related note, we were not happy about the fact that even long-time
supporters had to sign up again but we were not able to transfer
access from the old database. Sorry.
#2. Yes, we are asking people to email us for datasheets, software
packages and thumb drives (now called FORTHdrives) at this point and
this has to do with our desire to better understand what people are
trying to accomplish so we can improve and better support
applications. This email step is not a 'final' decision on my part
but I will admit that I don't think that it is wrong to ask for it
when you want something. The database has one simple purpose: to
allow us to inform people as soon as new software, hardware or updates
are available.
#3. Thumb drives (FORTHdrives): It is true; we don't have as many
thumb drives available as we had hoped for. We had unexpected board
quality issues that dropped the yield and added rework. So for the
time being we have to 'make them count', meaning that we indeed give
preference to people that have a chance of helping us create revenue.
To be sure, only few drives have gone to large companies! We are very
interested in people with great ideas! On a final and positive note,
we are optimistic that we will have significant quantities by end of
March to get around to most everybody who has approached us.
#4. As said, we are decidedly NOT ignoring the Forth community and/or
smaller players in order to cater to large companies only! We are,
however, trying to find a MANAGEABLE number of customers that will
bring in enough revenue to make us a viable and self-sustained
business at the earliest possible. A big concern is that our internal
support resources are very limited. If we get swamped with support
request without any revenue in sight it is a question of time until we
will have to throw the towel. Now, I realize that there are very
experienced and knowledgeable Forth programmers out there that might
never have to come back with a single support question. But how can I
tell if people aren't even willing to give me minimum information
before they grab our documentation and software?
#5. Distributors: I would love to have distributors promote our
chips, sell hardware (incl. FORTHdrives) and be able to handle all
kinds of support requests. Unfortunately I don't think that that is a
quick solution. As stated by Elizabeth, the SEAforth technology is
sufficiently different and complex enough to require significant
investments into training first (and we are planning that). I agree
that the community forum will eventually carry a lot of weight but is
for now still 'light in content'.
#6. Documentation: Documentation is now available and - while not
perfect - pretty substantial and comprehensive. This is not only done
for larger customers, full packages go out to essentially all parties
requesting the information. The reason it might still seem very
'light' is that we have only recently found time to create and
complete some important documents.
The bottom line:
We need to turn into a viable business or we will not be able to
continue. We would love to enable millions of engineers and even
hobbyists if we can be sure that the support activities are not
grounding us. I agree with some of the posts above that the
community forum has the potential to become an extremely powerful tool
in this context. For now, it is still too light and needs to
establish itself first. You all can help us in that mission.
Sorry for the long-winded post, I hope this helps explain things.
Jurgen
Howerd
Thank you for your post.
I can report that I received the SEAforth-24A datasheet shortly after
requesting it.
I have an application in mind, and I am hoping to get a couple of
chips ASAP.
Since there has been a lot of discussion in the past based on minimal
information, and there have been issues surrounding just what
information could or should be made public, it has been a real
pleasure to read the datasheet at last.
So here are my comments (please don't expect any pretense at
impartiality here) :
1. The datasheet is extremely clear and well written.
2. The level of detail is excellent - it appears to be a complete
description of the chip, right down to look-ahead in the instruction
decoder.
3. The chip itself is very well thought out : 24 bits of GPIO, two 6
bit DACs, two 18 bit ADCs, memory chip interface, various boot options
from EEPROM, FLASH or serial port. The two ADCs are based on two free-
running counters that can clock at up to around 25GHz.
4. The individual C18 processor cores are extremely elegant. Hardware
constraints mean that the opcode patterns in the 18 bit words are
complex, leading to a complex compiler. Of course, complex in this
context means maybe a Kbyte of code...
5. The AC timing data contains a lot of "TBD"s. Rumour has it that the
C18 cores clock at around 600MHz.
To sum up :
Chuck Moore has raised the bar in processor design, again.
The SEAforth-24A chip is to the Pentium what colorForth is to Vista.
Please pass on my best wishes and congratulations to the team on a job
well done!
Regards
Howerd 8^)
On 5 Feb, 21:08, SEAforth Marketing <seaforth.market...@gmail.com>
wrote:
Howerd
Sorry, I miscalculated the free-running counter speed - I believe they
are 6.5GHz, not 25GHz. Still fast enough, though. I will confirm this
when I get the real chips...
Regards
Howerd 8^)
m-coughlin
>
> All,
>
> Please allow me to jump in and confirm a couple of things said,
> defuse some rumors and add some facts. First, my name is
> Jurgen Krehnke and I am responsible for Marketing and Sales
> at IntellaSys.
>
You have the hardest job in the company. Forth is the best
way to program computers but only a few fanatical programmers
know this and they are unable to demonstrate its value. You
might produce a useful hardware product, but it is only a minor
improvement compared to the improvement Forth makes in programming.
> #2. Yes, we are asking people to email us for datasheets,
> software packages and thumb drives (now called FORTHdrives)
> at this point and this has to do with our desire to better
> understand what people are trying to accomplish so we can
> improve and better support applications. This email step is
> not a 'final' decision on my part but I will admit that
> I don't think that it is wrong to ask for it when you want
> something. The database has one simple purpose: to allow
> us to inform people as soon as new software, hardware or
> updates are available.
<snip>
> A big concern is that our internal support resources are
> very limited. If we get swamped with support request without
> any revenue in sight it is a question of time until we will
> have to throw the towel. <snip>
> But how can I tell if people aren't even willing to give me
> minimum information before they grab our documentation and
> software?
You've got things completely backwards. Your product
literature must be read, read widely, and be worth reading.
Worry when your resources are strained by support requests --
afterwards. First worry that your services will be requested.
Just being involved with Forth will discourage most of your
potential big customers. They will want to know how good is your
C compiler. You will have to spend much effort to educate them
differently. The idea of asking for something before you will
send out your technical data is extrodinately bad. You must get
large numbers of technical people to look at your documentation.
Only a few of your potential clients will take the trouble to
alter their thinking and realize they should use your products.
<snip>
> As stated by Elizabeth, the SEAforth technology is
> sufficiently different and complex enough to require significant
> investments into training first (and we are planning that).
<snip>
Planning that? PLANNING?? Training is FIRST. Its the FIRST
thing a company like yours needs to do. Getting the parts
working is the second. You are not doing the same old
programming thing. Chuck Moore says the software problem has
been solved. It is up to you, the Marketing and Sales director,
to show the whole computer industry what Chuck Moore is talking
about. Nobody has been able to do that for the past thirty or
fourty years. Make sure everybody in your company knows this and
that you are expecting their help in this matter.
> #6. Documentation: Documentation is now available and - while
> not perfect - pretty substantial and comprehensive. This is
> not only done for larger customers, full packages go out to
> essentially all parties requesting the information.
> The reason it might still seem very 'light' is that we have
> only recently found time to create and complete some important
> documents.
<snip>
> Jurgen
How can you be so wishy-washy about documentation? People
will want to learn to use your poduct because you have the best
documentation, not just because you have the best hardware.
Early in my career I was puzzled as to why companies like
Digital Equipment, Motorola and National Semiconductor would
give me all the data books I asked for, even more than I could
carry! Now I know it was because they wanted everybody in the
world who might be able to understand them to read them. The
fewer people who understand what they make, the fewer parts they
will sell. The Internet and world wide web eliminate the cost of
paper and printing. Take maximum advantage of that. Make
documentation number one, not number six.
--
Michael Coughlin m-cou...@comcast.net Cambridge, MA
Charlie Springer
(in article <13q44qe...@news.supernews.com>):
> Intellasys is trying to focus their sales and support efforts on
> potentially significant users of the technology (folks who will buy
> large quantities of parts). That's a common business model for
> companies introducing new and innovative technologies. Harris Semi
> followed a similar plan in the 80's with the RTX parts.
The PIC processors took the opposite aproach and filled the demand from
individuals, which has worked out pretty well. I assume Intellasys wants ARM7
sized design wins, but thaqt didn't happen till the, well, 7th generation of
ARM.
-- Charlie Springer
Jecel
> The PIC processors took the opposite aproach and filled the demand from
> individuals, which has worked out pretty well. I assume Intellasys wants ARM7
> sized design wins, but thaqt didn't happen till the, well, 7th generation of
> ARM.
Which is why Apple gave them the brilliant idea of skipping the ARM4
and ARM5 :-)
But I agree with what you and other people are saying. If I have to
look at the sites and datasheets for eight new processors to select
one for my next product, I won't bother with the ones that make it
harder for me. Intellasys is lucky that most of their current direct
competitors aren't much better (so far my worst experience has been
with Marvell and its silly NDAs for mere datasheets) but TI, NXP
(Philips) and others are happy to give me what I need.
-- Jecel
John Passaniti
> But I agree with what you and other people are saying. If I have to
> look at the sites and datasheets for eight new processors to select
> one for my next product, I won't bother with the ones that make it
It's actually a bit worse for Intellasys in that before engineers
looking at their chips can go through the bother of getting datasheets,
they have to first know about those chips. And unless they happen to be
hanging out in comp.lang.forth or seeing a typically glowing press
release with the usual superlatives, they don't even know about their
processors.
Here's the reality of how engineers start out finding out about chips:
Google. When I'm looking for a chip to solve a problem, I'm going to
toss a few keywords in Google, and then I'll sort through what I get
back. And far more often than not, my priority is to spend the most
time with any datasheets I find, not in sending email to the company and
asking if, pretty please, I can get datasheets.
Here's an example. I'm looking for a chip that does accurate
measurement of powerlines. So I toss a few important keywords at Google
("voltage measurement iRMS vRMS reactive apparent chip") and in the
first page, I get chips from Analog Devices and Cirrus Logic. In two
clicks, I have datasheets on my screen and I can read about them, and
more important than that, I know two manufacturers exist that have what
I want. I can then look at both manufacturers more carefully, or I can
expand my scope and see if there are any other competitive products.
But chances are that unless the chips I'm looking at are undesirable,
I'm going with them.
The point here is that because both Analog Devices and Cirrus Logic put
their datasheets online, they are searchable by Google and other search
engines. All they did was put a link to a datasheet on their website
and now it acts as a kind of passive marketing of the chips to people
who are actively searching for them.
Jerry Avins
There's a billboard visible from I-95 near Philadelphia that reads
DON'T ADVERTISE
You might run out of stock.
Jecel
> [agree that easy access to datasheet is important]
Today I went to the Intellasys site again and the datasheet is now
freely available for download, with lots of clear pointers to it. I
hope this move gets them more customers, as we had been discussing
here.
-- Jecel
Albert van der Horst
I saw it too. Kudos to them. Even more so if it means that they
took the good advice of Passiniti e.a. at heart.
[I took no chances though and downloaded it *immediately*.]
>
>-- Jecel
Groetjes Albert
--
--
Albert van der Horst, UTRECHT,THE NETHERLANDS
Economic growth -- like all pyramid schemes -- ultimately falters.
albert@spe&ar&c.xs4all.nl &=n http://home.hccnet.nl/a.w.m.van.der.horst
John Passaniti
> I saw it too. Kudos to them. Even more so if it means that they
> took the good advice of Passiniti e.a. at heart.
Or even Passaniti. I'm not one of those people who flips out if my last
name isn't spelled correctly, but I do note that it is only here in
comp.lang.forth where people seem consistently unable to copy-and-paste.
This isn't an issue in the various other forums I participate in.
In any case:
> [I took no chances though and downloaded it *immediately*.]
I did too, and I'm happy to see additional detail over the last data sheet.
Bruce McFarling
> Or even Passaniti. I'm not one of those people who flips out if my last
> name isn't spelled correctly, but I do note that it is only here in
> comp.lang.forth where people seem consistently unable to copy-and-paste.
> This isn't an issue in the various other forums I participate in.
True Forth enthusiasts always type it in from scratch.
John Passaniti
Then apparently, comp.lang.forth doesn't have many true Forth
enthusiasts because I am constantly delighted by the variety of ways
people spell my name here... and only here.
Bruce McFarling
> Bruce McFarling wrote:
> > True Forth enthusiasts always type it in from scratch.
> Then apparently, comp.lang.forth doesn't have many true Forth
> enthusiasts because I am constantly delighted by the variety of ways
> people spell my name here... and only here.
No, its all the mere copy and pasters over at all those *other* lists
that get it right ... its typing it in from scratch that introduces
all the *novelty*.
Charlie Springer
(in article
<8c9eeb57-899d-4ae7...@n58g2000hsf.googlegroups.com>):
> Which is why Apple gave them the brilliant idea of skipping the ARM4
> and ARM5 :-)
Hey! I have a VLSI ARM4 chip set.
-- Charlie Springer
Charlie Springer
(in article
<c707bdcf-6590-4f6b...@x41g2000hsb.googlegroups.com>):
> No, its all the mere copy and pasters over at all those *other* lists
> that get it right ... its typing it in from scratch that introduces
> all the *novelty*.
When I was first exposed to Forth with my AIm-65 I was so frustrated that I
called Rockwell. The problem? I couldn't spell proceeding or something like
that.
-- Charlie Springer
Jecel
> On Tue, 19 Feb 2008 19:07:11 -0800, Jecel wrote
> > and ARM5 :-)
>
> Hey! I have a VLSI ARM4 chip set.
You might want to recheck that.
- ARM1, 1985, was (mostly) never released outside Acorn.
- ARM2, 1986, was the first one sold by VLSI and was a faster ARM2
(newer process) with multiply. It also had a chip set.
- ARM3, 1989, was much faster and had a cache.
- ARM2aS, 1990, was a static version of the ARM2.
- ARM6, 1992, had a 32 bit PC mode and could be either big or little
endian
The VLSI VL86C410 was the i/o chip of the ARM2 chipset.
-- Jecel