Forum: GNU Radio Intel Atom is NICE.

Announcement (2017-05-07): www.ruby-forum.com is now read-only since I unfortunately do not have the time to support and maintain the forum any more. Please see rubyonrails.org/community and ruby-lang.org/en/community for other Rails- und Ruby-related community platforms.
E16be4811324adf8f26be26d77e9d29d?d=identicon&s=25 Bob McGwier (Guest)
on 2008-12-28 21:24
(Received via mailing list)
I am running GnuRadio, SDRMAX, and  PowerSDR on my new Intel ATOM 330
MiniITX motherboard.  I had to put a firewire card in the single PCI
slot.
The integrated intel graphics are nice and spiffy (glxgears is at 800
fps if
you turn off the desktop enhancements, no wiggly windows please).



At 96000 PowerSDR is running under 15% CPU.  The ATOM burns EIGHT watts
and
has dual hyperthread cores  (shows up as four processors in task
manager).



My intel ATOM computer, case, memory, disk, keyboard/mouse cost me $275.



The motherboard is available for $85  (with processor on it, air cooled)
at
NewEgg.   I used it for firewire since there is no firewire.  You no
longer
need the LCD display stuff so I hope that is dumped.

The ATOM 330 supports SSE,SSE2, SSE3, SSSE3 so it will run SIMD code
quite
nicely and does so with GnuRadio, SDRMAX, and PowerSDR.





Intel atom mobo/processor bundle: $85

http://www.newegg.com/Product/Product.aspx?Item=N8...



case:  $56

http://www.newegg.com/Product/Product.aspx?Item=N8...



memory:  $21

http://www.newegg.com/Product/Product.aspx?Item=N8...



drive:  $79



http://www.newegg.com/Product/Product.aspx?Item=N8...





The case has a completely overkill 250w power supply.  They must be
expecting you to put a Pentium D MiniITX board in there.  ;-).  It was
an
excessive thing anyway and done only because it will sit next to the
television in the den.



Already had keyboard and mouse and the firewire card was $12 the last
time I
looked.  The (new from Santa) 52" TV is my monitor.  Running Ubuntu 8.10
and
Windows XP Pro.    I have used it to do GnuRadio, PowerSDR, SDRMAX II,
streamed video's from Netflix, Hulu, Youtube, and more.



This is a cheap enough, high enough performance computer to complete
dedicate it to the task, and never run into a stupid outlook glitch for
PowerSDR.  On Ubuntu, it is spiffy to compile GnuRadio pretty quickly
with
make -j4.



I will give some GnuRadio and SDRMaxII numbers later.  Phil Covington
has
one so maybe he can give us SDRMax II numbers while I concentrate on
GnuRadio/DttSP.



I will be the first to admit it is not as fast as my QX6700 Ubuntu
machine
or the new I7 extreme machine (dual Fedora, Vista 64 for Nvidia Tesla
programming which requires a GOOD pciE-x16 slot and SLI support and a
$400
Mobo) but it is plenty fast for these dedicated job small CPU tasks
(SDR)
and really cheap.  I made no great effort to save money, it is just
cheap
anyway.



I am pushing ahead on the Beagleboard because  I think we need to
collectively turn some attention to the embedded low power SoC and I
combine
(along with my embedded programming friends) both ARM and SIMD DSP for
SDR.



Happy New Year,

Bob





ARRL SDR Working Group Chair

Member: ARRL, AMSAT, AMSAT-DL, TAPR, Packrats,

NJQRP, QRP ARCI, QCWA, FRC.

"And yes I said, yes I will Yes", Molly Bloom
C7587810780b7d714e062e93c6955868?d=identicon&s=25 Daniel O'Connor (Guest)
on 2008-12-29 00:12
(Received via mailing list)
On Monday 29 December 2008 06:53:30 Bob McGwier wrote:
> I am running GnuRadio, SDRMAX, and  PowerSDR on my new Intel ATOM 330
> MiniITX motherboard.  I had to put a firewire card in the single PCI slot.
> The integrated intel graphics are nice and spiffy (glxgears is at 800 fps
> if you turn off the desktop enhancements, no wiggly windows please).
>
>
>
> At 96000 PowerSDR is running under 15% CPU.  The ATOM burns EIGHT watts and
> has dual hyperthread cores  (shows up as four processors in task manager).

Unfortunately the 945 chipset eats ~20W - god knows why Intel lumbered
the
Atom with it :(

Toms Hardware (and others) did a test of the Atom vs an underclocked
Athlon
and the later won most of the tests

http://www.tomshardware.com/reviews/Atom-Athlon-Ef...

I think the Atom combo is cheaper and smaller though :)
E16be4811324adf8f26be26d77e9d29d?d=identicon&s=25 Bob McGwier (Guest)
on 2008-12-29 11:59
(Received via mailing list)
The intel graphics chip set and northbridge are power hungry.  I think
the idea is optimize code for the 330 and not the peripherals with this
ine

ARRL SDR Working Group Chair
Member: ARRL, AMSAT, AMSAT-DL, TAPR, Packrats,
NJQRP, QRP ARCI, QCWA, FRC.
"And yes I said, yes I will Yes", Molly Bloom
E16be4811324adf8f26be26d77e9d29d?d=identicon&s=25 Bob McGwier (Guest)
on 2008-12-29 12:11
(Received via mailing list)
The intel graphics chip set and northbridge are power hungry.  I think
the idea is optimize code for the 330 and not the peripherals with this
inexpensive and easy to use Mobo.  The disk drive is hungry as well.

The Intel ATOM Z500 family are mobile processors that have the same SIMD
registers,  support LCD,  good GigE support, and the potential wart
(POTENTIAL) is the 100 or 133 MHz FSB.

http://www.axiomtek.com.tw/Products/ViewProduct.as...


The Virginia Tech Mobile (SDR/CR) groups are looking at these (largest
SDR/CR department in the world I think).

I think we want to optimize SIMD code and see what we can get running on
these prepackaged, easy to get up and running systems as well as the SoC
parts such as that family of TI OMAP parts carried on the Beagleboard.

One does NOT need to spend thousands of dollars on an SDR computer for
most operations.   That is a convenient excuse for me to justify my
computer budget for "development" and get high end things to play with
(so I can kill aliens from another galaxy or FSU laboratory with
impressive graphics in my "spare time").  But most people need to really
justify the need for the high end computer in my mind and they cannot.
That is my point in all of this.

SDR wants to be on consumer/commodity level processors and SoC to be in
everyone's "coffee budget" and taken for granted in the ideal world in
my book.  There seems to be little gained by optimizing this for Quad
core extreme processors with massive GPU's sitting on them,  tons of
expensive high speed memory, and the world's fastest drives and costing
well upwards of $1000 US.  Lightweight,  easy to distribute, with
browser level GUI's and distributed everything on inexpensive processors
and we rule the world.  You can call me Dr. No. and SDR Working group
chairman stands for SPECTRE DUMB RESEARCH working group.

Cheers,
Bob


ARRL SDR Working Group Chair
Member: ARRL, AMSAT, AMSAT-DL, TAPR, Packrats,
NJQRP, QRP ARCI, QCWA, FRC.
"And yes I said, yes I will Yes", Molly Bloom
C7587810780b7d714e062e93c6955868?d=identicon&s=25 Daniel O'Connor (Guest)
on 2008-12-29 12:53
(Received via mailing list)
On Monday 29 December 2008 21:41:03 Bob McGwier wrote:
> The intel graphics chip set and northbridge are power hungry.  I think the
> idea is optimize code for the 330 and not the peripherals with this

I think that unfortunately the power consumption is high regardless of
wether
you actually use the video or not :(

> inexpensive and easy to use Mobo.  The disk drive is hungry as well.

Well you could use an SSD ;)

> One does NOT need to spend thousands of dollars on an SDR computer for most
> operations.   That is a convenient excuse for me to justify my computer

I think the Athlon would be quite competitive, it has higher memory
bandwidth
I believe and the boards aren't limited in connectivity like the Atom.

Basically my point was that it should be considered rather than just
assuming
the Atom is best :)
558c40b97bd1af8d912424757714bda9?d=identicon&s=25 Marcus D. Leech (Guest)
on 2008-12-29 14:14
(Received via mailing list)
Bob McGwier wrote:
> One does NOT need to spend thousands of dollars on an SDR computer for most operations. 
That is a convenient excuse for me to justify my computer budget for "development" and get 
high end things to play with (so I can kill aliens from another galaxy or FSU laboratory 
with impressive graphics in my "spare time").  But most people need to really justify the 
need for the high end computer in my mind and they cannot.  That is my point in all of 
this.
>
> SDR wants to be on consumer/commodity level processors and SoC to be in everyone's 
"coffee budget" and taken for granted in the ideal world in my book.  There seems to be 
little gained by optimizing this for Quad core extreme processors with massive GPU's 
sitting on them,  tons of expensive high speed memory, and the world's fastest drives and 
costing well upwards of $1000 US.  Lightweight,  easy to distribute, with browser level 
GUI's and distributed everything on inexpensive processors and we rule the world.  You can 
call me Dr. No. and SDR Working group chairman stands for SPECTRE DUMB RESEARCH working 
group.
>
>
While I can heartily agree that for the expansion of SDR into the
consumer space, you want it to run on low-power processors, etc, I can't
  agree that "for most operations" you don't need a high-end CPU.

For example, 802.11 at anywhere approaching 802.11b bitrates needs some
serious iron, and yet in our world (the world of SDR
  geeks), wanting to build SDR/GnuRadio-based 802.11b implementations
seems a fairly common goal.

In my work in radio astronomy, I've found that despite the relative
simplicity of the basic functions my software provides--full-bandwidth
  spectral display, and total power, for one or two channels, big iron
is necessary.   I recently upgraded to a quad-core Q6600 to replace
  a dual-core Pentium D 940.  The quad core loses against the dual-core
because of a difference in maximum clock speed.  I can
  run the D 940 at 3.2Ghz forever, and it can process a full 8Mhz of
dual-channel, complex bandwidth.  The Q6600, on the other hand,
  is unstable above 2.85Ghz or so, and can't sustain more than about
5.3Mhz of dual-complex-channel bandwidth without incurring
  massive USRP overruns.

Despite the wonderful new multi-threaded Gnu Radio framework, it seems
that at least one of those threads really needs as many
  MIPs as the processor can throw at it, because it has to keep up with
a real-time data source.

Any time you're dealing with having to suck in (or send out) as much
bandwidth as the USRP can tolerate, and
  *actually doing something* with the entire bandwidth, you need
ManyMIPS(tm).
  Which means spending $$$ (although, my dual/quad-core system was much
less than the $1000.00 you quote above).

--
Marcus Leech
Principal Investigator, Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org
6b9e86b84231f94e35327d5747111bcd?d=identicon&s=25 Newman, Timothy (Guest)
on 2008-12-29 16:06
(Received via mailing list)
There are many more ways than just lumping everything onto a single GPP.
A good example is a recent thread on the GNU radio mailing list where
the poster is using the USRP2 as a standalone radio with no PC.  Pushing
key elements to other reconfigurable processors, e.g. the USRP2 FPGA,
will greatly ease the burden of the GPP.  My point is that "big iron"
isn't always necessary if you're willing to put some work into
distributing the work load to other processors (a major research issue
currently).

Tim


> -----Original Message-----
> From: discuss-gnuradio-bounces+trnewman=vt.edu@gnu.org [mailto:discuss-gnuradio-
> bounces+trnewman=vt.edu@gnu.org] On Behalf Of Marcus D. Leech
> Sent: Monday, December 29, 2008 8:14 AM
> To: Bob McGwier
> Cc: hpsdr@lists.hpsdr.org; qs1r@yahoogroups.com; FlexRadio@flex-radio.biz; discuss-
> gnuradio@gnu.org
> Subject: Re: [Discuss-gnuradio] Intel Atom is NICE.
E16be4811324adf8f26be26d77e9d29d?d=identicon&s=25 Bob McGwier (Guest)
on 2008-12-30 01:40
(Received via mailing list)
And GPU's are going to become commodity priced quickly and possibly even
move into the GPP and replace older ways of doing floating point.  With
Nvidia CUDA, you can write code for your GPP, call GPU with intrinsics
to get pretty quick payback while a better longer term strategy is
worked on.

The future of really hard to program heterogeneous/not symmetric
multiple core processors,  irrespective of how great the bandwidth is,
I don't think is looking all that rosy.  It simply cannot take months
and months to get speed to make the processor pay or the cost per flop,
when ALL COSTS are amortized (expensive people, etc.) begins to look
bad.

Bob



ARRL SDR Working Group Chair
Member: ARRL, AMSAT, AMSAT-DL, TAPR, Packrats,
NJQRP, QRP ARCI, QCWA, FRC.
"And yes I said, yes I will Yes", Molly Bloom
Cf1f87de124043f2da6c95cfc293358e?d=identicon&s=25 Eric A. Cottrell (Guest)
on 2008-12-30 04:54
(Received via mailing list)
Bob McGwier wrote:
> And GPU's are going to become commodity priced quickly and possibly even move into the 
GPP and replace older ways of doing floating point.  With Nvidia CUDA, you can write code 
for your GPP, call GPU with intrinsics to get pretty quick payback while a better longer 
term strategy is worked on.
>
> The future of really hard to program heterogeneous/not symmetric multiple core 
processors,  irrespective of how great the bandwidth is,  I don't think is looking all 
that rosy.  It simply cannot take months and months to get speed to make the processor pay 
or the cost per flop, when ALL COSTS are amortized (expensive people, etc.) begins to look 
bad.
>
> Bob
>
Hello,

I bought the earlier version of the motherboard with just the 10/100
ethernet.  I put it in a MI-100 case and it is a nice little system.  I
have not gotten a chance to use it with GNURadio much so I have not
commented about it on the list.  I am thinking of using it as a car
computer.

However, I have tried it under Windows XP with my AutumnWave OnAir GT
USB HDTV tuner.  I understand this tuner uses the computer to decode the
HDTV signal.  On my older Thinkpad z60t (1.73 GHz Pentium M) the HDTV
decode takes about 80% CPU and there are ocassional gliches.  I was not
happy.  With the Intel Atom (1.60 GHz), the HDTV decode takes about 40%
CPU and decode is excellent.  I suspect the hyperthreading feature of
the CPU is causing the improved decoding.

I recently bought a Acer Aspire One (Intel Atom and 120GB HD) on sale.
I plan to put Linux on it and see how it works with GNURadio and the
USRP.  I feel the advantage of the Intel Atom is reasonable performance
in a small package.  The Atom-based netbook may be useful for dealing
with ham and conventional narrow-bandwidth signals in a portable/mobile
SDR package.

The 945 chipset is power hungry but I heard there is another chipset for
the Atom in the works.  Hopefully the new chipset will allow using the
GPU for SDR work.

73 Eric
Cf1f87de124043f2da6c95cfc293358e?d=identicon&s=25 Eric Cottrell (Guest)
on 2009-01-05 04:28
(Received via mailing list)
----- Start Original Message -----
Sent: Mon, 29 Dec 2008 22:53:06 -0500
From: "Eric A. Cottrell" <wb1hbu@runbox.com>
To: discuss-gnuradio@gnu.org
Subject: Re: [Discuss-gnuradio] Intel Atom is NICE.

> Hello,
>
> I bought the earlier version of the motherboard with just the 10/100
> ethernet.  I put it in a MI-100 case and it is a nice little system.  I
> have not gotten a chance to use it with GNURadio much so I have not
> commented about it on the list.  I am thinking of using it as a car
> computer.

----- End Original Message -----

Hello,

I did not realize that the D945GCLF2 has the new Atom 330 dual core
processor. It should work even better than the earlier D945GCLF board or
netbook that I used.

73 Eric
This topic is locked and can not be replied to.