[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [oc] Jim Turley's article at Embedded.com
Heya !
I'm thinking what is the fundamental difference between open source (free)
software and open source hardware (particulary IP cores). I guess the most
fundamental difference is in the user and developer base.
For development of an open source software you need a person with some
programming skills and time, a computer and some software development tools.
Everything is available for a long time.
For development of an open source IP core you need a person with some HDL
coding skills and time, a computer, some EDA software tools and means of
transferring the IP core into hardware (last one is not a must but is
helps). A person and a computer is not a problem. *Good* set of open source,
free EDA tools doesn't yet exist and *good* commercial tools are still quite
expensive. Transferring an IP core into HW is even more expensive, since a
decent FPGA development board can cost easily several thousand USD (decent
means an FPGA board that doesn't limit you all the time with area/speed). An
ASIC is probably outside of individual's reach (between $50k and $1M
depending what technology (and yes I know MPW shuttle runs Mosis,
Europractice etc...)).
Even more important. The User ! In case of OSS/FS the user is simply a
person with a computer and enough brain cells to know how to compile or
simply install a binary. With over 300M PCs out there in the world,
potential user base for OSS/FS is huge.
Who is the user base for open source Ip cores? Definately not 300M PC users
since there is no FPGA in their PC and they have no means to create an ASIC.
So who are our potential users. Companies, universities, companies,
companies. Until either Intel or AMD doesn't put an FPGA in their x86 ...
So I'm thinking what can we do to attract companies and universities (and I
don't mean all those students that need to do their "final" project). Some
ideas that would help expose our work, get more developers hooked and - most
importantly - create user base:
1) Get companies/universities to form partnerships with us (us == opencores)
2) Go to exhibits and trade shows?
3) Write articles for magazines ?
4) Send a newsletter that would help expose our work?
5) Allow companies to put ads on opencores site whenever they use an IP core
in their product (this translates into benefits for both a company and gives
a reference for our IP core)?
6) Form a computing farm with all the cool EDA tools (simulators, linters,
synthesis ...), possibly also connect development boards (more for ATS
(automatic testing system - a form of verification) than anything else)
7) what else ...?
I think we will have to follow companies' models and create a marketing and
sales teams or won't get very far. ;-))
regards,
Damjan
----- Original Message -----
From: "Jerrold Wen" <jwen@visualsonics.com>
To: <cores@opencores.org>
Cc: <jim@jimturley.com>
Sent: Tuesday, June 04, 2002 4:09 PM
Subject: Re: [oc] Jim Turley's article at Embedded.com
> <RANT ON>
>
> Sorry, I just finished reading this article, and I couldn't help but feel
angry
> at the ignorance and narrow-minded point of view that this guy gives. I
mean
> the title of the article is Open-Source Hardware, but it clearly does not
even
> touch on the topic. It only considers open-source hardware as 'free
> microprocessor', ie. Intel. Clearly, we are a long way off from the
equivalent
> of Linux to Windows on the microprocessor side. But there are many other
> places where open-source hardware IP is advantageous - such as
controllers,
> glue logic, interfaces, memory, etc. There is not a single mention of
> reconfigurable, reprogrammable logic anywhere. Nor of hardware
development
> systems or platforms.
> And these arguments of hardware not being tweakable or easy-to-collaborate
are
> clearly fantasy opinions from somebody who no understanding of HDL code or
> development. Why, all these big semi companies must have only one
super-geek
> working on each chip! how do they do it?! without "modular and
> compartmentalized" code and code re-use....
> And the issue of no support?! Why would open-source hardware have any
more or
> less support than open-source software? Afterall, in the end it's still a
> group of developers who have either a good, moderate, or no understanding
of a
> particular piece of code.
> The author should of spent just 5 mins more of research to find out what
open-
> source hardware really means before he disseminated this garbage and
obviously
> negative opinion. A quick search would have found groups like
opencores.org or
> openh.org.
>
> Free your mind,
> Jerrold.
>
> Quoting Pablo Bleyer Kocik <pbleyer@embedded.cl>:
>
> >
> > Embedded.com has just published an "interesting" article about
> > Open-Source Hardware by Jim Turley. Too biased... if you ask me.
> >
> > http://www.embedded.com/printableArticle?doc_id=OEG20020524S0078
> >
> > Cheers!
> > --
> > PabloBleyerKocik/
> > pbleyer /"Rintrah roars & shakes his fires in the burdend air;
> > @embedded.cl/ Hungry clouds swag on the deep" - William Blake
> >
> > --
> > To unsubscribe from cores mailing list please visit
> > http://www.opencores.org/mailinglists.shtml
> >
> --
> To unsubscribe from cores mailing list please visit
http://www.opencores.org/mailinglists.shtml
>
--
To unsubscribe from cores mailing list please visit http://www.opencores.org/mailinglists.shtml