[Top][All Lists]

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

Re: [Discuss-gnuradio] Minimal Hardware to Run Gnuradio on?

From: David Carr
Subject: Re: [Discuss-gnuradio] Minimal Hardware to Run Gnuradio on?
Date: Thu, 16 Dec 2004 01:55:49 -0600
User-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.3) Gecko/20040928 MultiZilla/


I don't see whats wrong with that at all. As long as your disks can handle the transfer rate you're set. Even for really high rate USB transfers (40MB/sec) I see around 10% utilization on a P4 2.4. Also, some of my initial research actually streamed the captured data from the capture machine over a network to other more gutsy machines. I don't know if that's an option in you application.

-David Carr

Tom Huppi wrote:

On Wed, 15 Dec 2004, David Bengtson wrote:

What are the minimum hardware requirement for Gnu Radio? I'm thinking in
terms of a Arm7/Thumb controller. If you try to build a portable thing
with Gnu Radio, it would be good to have something smaller than a laptop
that it runs on

I can't answer your question, but I think it's probably safe to
say that it depends a great deal on what one is trying to do, and
it seems that there are a lot of possibilities with a 'system' as
general as GnuRadio.

Lately I've been looking to upgrade my hardware with an eye toward
making use of GnuRadio for as much as may become possible.  I'd
be interested in hearing anyones opinions about whether my ideas
are nuts or not.

I've become interested in the mini-ITX form factor (having sworn
never to touch a laptop again.)  Via has just started to release a
chipset (cn400) which looks OK on paper.  I expect it to support
USB2 well, GigE, and SATA with hardware RAID0 all through the
southbridge.  Their CPU's have always been lame, and I sense that
this is the most important thing to many potential GnuRadio
applications.  Here's my thoughts:

- Stream all data from the USB to disk.  500G should hold hours
worth by my back of the envelope calcs in worst case, and maybe
the lame CPU would be powerful enough to do some initial crunching
to extend this.

- Have the potential to process the data with 0->n higher powered
machines over the network, and at a time of one's choosing.

Does this paradigm strike anyone as potentially possible,
assuming enough development work eventually occurs?

BTW, here's a diagram of Via's chipset:



- Tom

Discuss-gnuradio mailing list

reply via email to

[Prev in Thread] Current Thread [Next in Thread]