[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [f-cpu] Re: Project short description

On Mon, 3 Sep 2001, Michael Riepe wrote:

> On Mon, Sep 03, 2001 at 09:10:27AM +0200, Juergen Goeritz wrote:
> [...]
> > Long story... To make it short some headlines only -
> > new architecture for cluster computing - want to go for
> > EU or german funding for this project - performant user
> > processor still to be decided on (may be either a LEON
> > extended to 64bit or F-CPU or other 64bit architecture
> > that can be used for embedded) - system on chip design
> > already specified - team to be built for realization -
> > drawback: new architecture is not built as open source...
> If you make a SoC that includes an F-CPU core, the source code for the
> entire chip must be released under the terms of the GPL because it's a
> derived work.  If you can live with that restriction...

No, I can't. I was wondering if it could be changed to the
more open LGPL license style like with LEON. Is this possible?

On the other hand, if it's not possible to change to LGPL,
I would not want to invest my own money but only go for it
when there is enough funding money and goods. Up to now my
opinion was that you can't really live from open projects.
But you are invited to proove the opposite...

> > I like f-cpu because it is extendable to higher bitwidth
> > later on and seems to be designed for low gate count.
> *giggle* didn't you see the multiplier?

:-) multipliers always come with high gate count...

> > So what's the basic feeling of you people involved in
> > f-cpu design and simulation to include f-cpu into a
> > system? I am interested in your views very much.
> Do you want only the design for your project, or the design team as well?

Good question. If I had a team it would be much easier
to get it done. On the other hand I don't have enough
money to set it up just for fun. I want to build a team
at Karlsruhe or Baden-Baden (Germany) to focus on my
mCluster project. But the people inside this team would
be free to work on f-cpu as well since it may be a part
of the whole project. Other companies and universities
are interested and will be invited to participate as well.

> You can do with the design whatever you like, as long as you follow
> the rules.  But isn't it a bit too early?  We're still in the first of
> many implementation phases, and we can give no guarantees that the F-CPU
> will be finished some day, that it will perform well, or work at all.
> If you rely on that, your project may fail.

There may be the fallback to use an extended LEON design
in the first place. Don't know if you like that. It's a
big challenge anyway...


To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/