Loading ...
Sorry, an error occurred while loading the content.
 

Re: [sdr-radio-com] Re: Multiple SDRs at once

Expand Messages
  • Ken N9VV
    Joe, take a look at the work that Hermann DL3HVH is doing with OpenCL. It will make use of multiple cores, hyper-threads, and the GPU available in the Nvidia
    Message 1 of 25 , Aug 3, 2013
      Joe, take a look at the work that Hermann DL3HVH is doing with OpenCL.
      It will make use of multiple cores, hyper-threads, and the GPU available
      in the Nvidia cards. He is coding SDR software called "cuSDR64" and it
      will out certainly outperform anything you have tested so far.
      Link: <URL: http://www.n9vv.com/N9VV-OpenCL.html >
      <URL: https://plus.google.com/107168125384405552048/posts >

      what a wonderful time to be alive and involved with SDR!
      rgds,
      Ken

      On 8/2/2013 9:36 PM, Joe Rotello wrote:
      > On 8/2/2013 9:37 PM, Ken N9VV (Win-7/64) wrote:
      >> Forgive me Joe, but you may be totally wrong in your assessment. There
      >> is a part of the Operating System called the "scheduler". In efficient
      >> Operating Systems, the scheduler (there may be many available as in
      >> Linux) tends to try and meet specific criteria -- some try to be fair
      >> and balanced, others like Windoze generate huge backlogs of DSP
      >> (Delayed Procedure Calls) because the Scheduler is so poorly written.
      >>
      >> If you grab some simple diagnostic tools like this Latency Monitor
      >> (DPC display), it will show you where your bottle neck actually is in
      >> your particular system:
      >> <URL: http://www.resplendence.com/download/LatencyMon.exe
      >> >
      >>
      >> there are many others that are also helpful. Getting OFF windows just
      >> for some tests on a bootable Linux USB stick, may show you an
      >> astonishing performance increase in old dual-core equipment :-)
      >>
      >> GL de Ken N9VV
      >
      > Sorry, my comments come from already testing the ideas of multiple-SDR's
      > running, and the severe loads on any and all GDI type OS's show that
      > Windows, Mac and Linux-based hardware all have the same "troubles" when
      > running multiple applications like SDR's that especially extract and
      > massage multiple FFT's (fast Fourier transformations) that lie at the
      > heart of most all current technology SDR's.
      >
      > Try it if you must, but the writing is already on the wall.
      >
      > Also, other SDR designers and makers have tried what you are wanting to
      > do, and their results so far have been the same as I am explaining.
      >
      > HOWEVER, that said...I am a champion to DO what you are wanting to do!
      >
      > SO one idea being toyed with is take the same SDR IC chip track as done
      > with multiple-core CPU's on PC's and laptops...build SDR chip-sets like
      > those that are about to appear, where the USB dongle will combine the
      > SDR and the up-converter. in one USB dongle, as opposed to the likes of
      > what we have right now...separate USB dongle being fed by a separate
      > up-converter board.
      >
      > The idea being discussed now is that instead of one SDR on the IC die,
      > vertical structuring a second (or third) die structure so that you COULD
      > have and run multiple SDR's. The dies would allow offloading as much of
      > the host systems OS loads as possible.
      >
      > Really, I am all for what you are wanting, but until you delve deeply
      > into how CPU's work in today's pre optical computer, pre-quantum
      > computer world, you may be sorely disappointed as of right now.
      >
      > Joe
      >
      >>
      >>
      >> On 8/2/2013 8:29 PM, josephrot wrote:
      >>> Hmm, could be.
      >>>
      >>> Keep in mind that the operational speeds would depend on oines CPU
      >>> speeds, and how many cores ones CPU has.
      >>>
      >>> Would think that for multiple SDR's all doing their own FFT's, the load
      >>> would "kill" a single-core CPU load-wise...
      >>>
      >>> ...but IF the software can recognize and support multiple-cor CPU'S,
      >>> that might work with a dual/triple/quad core Intel or AMD CPU box of
      >>> sufficient CPU speeds and of course usable memory.
      >>>
      >>> Still, the software would have to be very likely machine language to
      >>> keep up with the demands of say three or four SDR's at once.
      >>>
      >>> Joe
      >>>
      >>> --- In sdr-radio-com@yahoogroups.com
      >>> <mailto:sdr-radio-com%40yahoogroups.com>, "k6ozy" <yahoo@...> wrote:
      >>> >
      >>> > I was wondering if it is possible or planned to use multiple dongles
      >>> in the software at once. I can imagine how cool it would be if I had 4
      >>> or so RTL dongles attached, and then each one provides 2.5mhz of
      >>> spectrum placeable and moveable anywhere. The dropped VFOs would use any
      >>> dongle that happens to have been scrubbed under that VFO's location.
      >>> >
      >>> > Thoughts?
      >>> >
      >>>
      >>>
      >
      >

      --
      +°´°+,¸¸,+°´°~ ~°´°+,¸¸,+°´°+
    • josephrot
      Yes, I already have spoken in the past with Hermann on this. The use of the Nvidia GPU is something that generally works, and in some cases, one had to wait
      Message 2 of 25 , Aug 3, 2013
        Yes, I already have spoken in the past with Hermann on this.

        The use of the Nvidia GPU is something that generally works, and in some cases, one had to wait until Nvidia CUDA was built more into their products. Ditto for the AMD APP GPU's as well, although I favor Nvidia a LOT since I "grew up" using and recommending their video products and discrete chip-sets.

        The video card and/or chipset GPU (Graphics Processing Unit) can be used for FFT, can be used for both audio and video purposes, et al.

        The "trouble" is that the "GPU code" is generally having to be written for the specific maker of the GPU, but this is not always the case, and in fact this is less and less of a headache. There are some very good standards in place, and the GPU's are getting almost as "powerful" as the CPU's that they are taking the processing load off of.

        Yes, indeed...there's a LOT to be thankful for and to be expected almost on a weekly basis now, in SDR...

        And you are right on...it's a great time to be alive and at least interested in SDR.

        Joe

        --- In sdr-radio-com@yahoogroups.com, Ken N9VV <n9vv@...> wrote:
        >
        > Joe, take a look at the work that Hermann DL3HVH is doing with OpenCL.
        > It will make use of multiple cores, hyper-threads, and the GPU available
        > in the Nvidia cards. He is coding SDR software called "cuSDR64" and it
        > will out certainly outperform anything you have tested so far.
        > Link: <URL: http://www.n9vv.com/N9VV-OpenCL.html >
        > <URL: https://plus.google.com/107168125384405552048/posts >
        >
        > what a wonderful time to be alive and involved with SDR!
        > rgds,
        > Ken
        >
        > ( snipped for brevity )
      • Leif Asbrink
        Hi Joe and all, I found some statements below a bit provocative. I think modern computers have all the processing power we need. I have tested 7 RTL-SDR
        Message 3 of 25 , Aug 5, 2013
          Hi Joe and all,

          I found some statements below a bit provocative. I think modern
          computers have all the processing power we need.

          I have tested 7 RTL-SDR dongles running simultaneously with 7 instances
          of Linrad at the maximum sampling rate of 2.4 MHz on a 8-core machine.
          It is not the latest generation, but the CPU load is 43% only and there
          is a lot of resources available for decoders of various kinds. Operation
          seems totally stable and I have no reason to believe there would be any problem
          to run a couple of dozen of decoders.

          For unknown reasons Windows is not as efficient. Running with the same
          parameters under Windows 7 causes a CPU load of 90%.
          Operation is stable, but there is not much resources left for decoders.

          Note that Linrad runs a full FFT with waterfall for each one of the dongles
          and computes a narrowband signal for a selected frequency in each case.

          Surely a single core would not do the work, but if the sampling rate
          were set to 240 kHz a Pentium4 would manage many many dongles:-)

          This URL:
          http://www.sm5bsz.com/lir/sdrcmp/bdr/bdr-1.htm
          shows how 5 different hardware are compared in a single computer
          with 5 instances of Linrad running at sampling rates of 96 or 100 kHz.
          The CPU load is 4%!!!

          73

          Leif / SM5BSZ




          On Fri, 02 Aug 2013 22:36:38 -0400
          Joe Rotello <joerotello@...> wrote:

          > On 8/2/2013 9:37 PM, Ken N9VV (Win-7/64) wrote:
          > > Forgive me Joe, but you may be totally wrong in your assessment. There
          > > is a part of the Operating System called the "scheduler". In efficient
          > > Operating Systems, the scheduler (there may be many available as in
          > > Linux) tends to try and meet specific criteria -- some try to be fair
          > > and balanced, others like Windoze generate huge backlogs of DSP
          > > (Delayed Procedure Calls) because the Scheduler is so poorly written.
          > >
          > > If you grab some simple diagnostic tools like this Latency Monitor
          > > (DPC display), it will show you where your bottle neck actually is in
          > > your particular system:
          > > <URL: http://www.resplendence.com/download/LatencyMon.exe
          > > >
          > >
          > > there are many others that are also helpful. Getting OFF windows just
          > > for some tests on a bootable Linux USB stick, may show you an
          > > astonishing performance increase in old dual-core equipment :-)
          > >
          > > GL de Ken N9VV
          >
          > Sorry, my comments come from already testing the ideas of multiple-SDR's
          > running, and the severe loads on any and all GDI type OS's show that
          > Windows, Mac and Linux-based hardware all have the same "troubles" when
          > running multiple applications like SDR's that especially extract and
          > massage multiple FFT's (fast Fourier transformations) that lie at the
          > heart of most all current technology SDR's.
          >
          > Try it if you must, but the writing is already on the wall.
          >
          > Also, other SDR designers and makers have tried what you are wanting to
          > do, and their results so far have been the same as I am explaining.
          >
          > HOWEVER, that said...I am a champion to DO what you are wanting to do!
          >
          > SO one idea being toyed with is take the same SDR IC chip track as done
          > with multiple-core CPU's on PC's and laptops...build SDR chip-sets like
          > those that are about to appear, where the USB dongle will combine the
          > SDR and the up-converter. in one USB dongle, as opposed to the likes of
          > what we have right now...separate USB dongle being fed by a separate
          > up-converter board.
          >
          > The idea being discussed now is that instead of one SDR on the IC die,
          > vertical structuring a second (or third) die structure so that you COULD
          > have and run multiple SDR's. The dies would allow offloading as much of
          > the host systems OS loads as possible.
          >
          > Really, I am all for what you are wanting, but until you delve deeply
          > into how CPU's work in today's pre optical computer, pre-quantum
          > computer world, you may be sorely disappointed as of right now.
          >
          > Joe
          >
          > >
          > >
          > > On 8/2/2013 8:29 PM, josephrot wrote:
          > >> Hmm, could be.
          > >>
          > >> Keep in mind that the operational speeds would depend on oines CPU
          > >> speeds, and how many cores ones CPU has.
          > >>
          > >> Would think that for multiple SDR's all doing their own FFT's, the load
          > >> would "kill" a single-core CPU load-wise...
          > >>
          > >> ...but IF the software can recognize and support multiple-cor CPU'S,
          > >> that might work with a dual/triple/quad core Intel or AMD CPU box of
          > >> sufficient CPU speeds and of course usable memory.
          > >>
          > >> Still, the software would have to be very likely machine language to
          > >> keep up with the demands of say three or four SDR's at once.
          > >>
          > >> Joe
          > >>
          > >> --- In sdr-radio-com@yahoogroups.com
          > >> <mailto:sdr-radio-com%40yahoogroups.com>, "k6ozy" <yahoo@...> wrote:
          > >> >
          > >> > I was wondering if it is possible or planned to use multiple dongles
          > >> in the software at once. I can imagine how cool it would be if I had 4
          > >> or so RTL dongles attached, and then each one provides 2.5mhz of
          > >> spectrum placeable and moveable anywhere. The dropped VFOs would use any
          > >> dongle that happens to have been scrubbed under that VFO's location.
          > >> >
          > >> > Thoughts?
          > >> >
          > >>
          > >>
          >
        • josephrot
          Leif...Thank you for your comments... OH, to be sure, your tests are valid...no argument there, really, HOWEVER...note how many cores you are running and the
          Message 4 of 25 , Aug 5, 2013
            Leif...Thank you for your comments...

            OH, to be sure, your tests are valid...no argument there, really, HOWEVER...note how many cores you are running and the extreme performance one might need -- can everyone go out and buy 2, 4 or even 8 core hardware, and then run say 6 to 7 instances of SDR ?

            My discussion was more just reminding people to note their performance needs when we may want to run multiple SDR on one machine, but yes, your advisory is welcome news -- multiple SDR can be done.

            Joe

            --- In sdr-radio-com@yahoogroups.com, Leif Asbrink <leif@...> wrote:
            >
            > Hi Joe and all,
            >
            > I found some statements below a bit provocative. I think modern
            > computers have all the processing power we need.
            >
            > I have tested 7 RTL-SDR dongles running simultaneously with 7 instances
            > of Linrad at the maximum sampling rate of 2.4 MHz on a 8-core machine.
            > It is not the latest generation, but the CPU load is 43% only and there
            > is a lot of resources available for decoders of various kinds. Operation
            > seems totally stable and I have no reason to believe there would be any problem
            > to run a couple of dozen of decoders.
            >
            > For unknown reasons Windows is not as efficient. Running with the same
            > parameters under Windows 7 causes a CPU load of 90%.
            > Operation is stable, but there is not much resources left for decoders.
            >
            > Note that Linrad runs a full FFT with waterfall for each one of the dongles
            > and computes a narrowband signal for a selected frequency in each case.
            >
            > Surely a single core would not do the work, but if the sampling rate
            > were set to 240 kHz a Pentium4 would manage many many dongles:-)
            >
            > This URL:
            > http://www.sm5bsz.com/lir/sdrcmp/bdr/bdr-1.htm
            > shows how 5 different hardware are compared in a single computer
            > with 5 instances of Linrad running at sampling rates of 96 or 100 kHz.
            > The CPU load is 4%!!!
            >
            > 73
            >
            > Leif / SM5BSZ
          • Simon GD4ELI/HB9DRV
            Leif, What do you think about the hackRF & bladeRF? I m looking forward to the bladeRF as a possible VHF / UHF / SHF satellite receiver. Simon GD4ELI/HB9DRV
            Message 5 of 25 , Aug 6, 2013

              Leif,

               

              What do you think about the hackRF & bladeRF? I’m looking forward to the bladeRF as a possible VHF / UHF / SHF satellite receiver.

               

              Simon GD4ELI/HB9DRV

              http://v2.sdr-radio.com/

               

              From: sdr-radio-com@yahoogroups.com [mailto:sdr-radio-com@yahoogroups.com] On Behalf Of Leif Asbrink


              I have tested 7 RTL-SDR dongles running simultaneously with 7 instances
              of Linrad at the maximum sampling rate of 2.4 MHz on a 8-core machine.
              It is not the latest generation, but the CPU load is 43% only and there
              is a lot of resources available for decoders of various kinds. Operation
              seems totally stable and I have no reason to believe there would be any problem
              to run a couple of dozen of decoders.

            • Simon GD4ELI/HB9DRV
              Joe, It can be done, but the bladeRF solution may well be easier and is certainly a better SDR receiver. Simon GD4ELI/HB9DRV http://v2.sdr-radio.com/ From:
              Message 6 of 25 , Aug 6, 2013

                Joe,

                 

                It can be done, but the bladeRF solution may well be easier and is certainly a better SDR receiver.

                 

                Simon GD4ELI/HB9DRV

                http://v2.sdr-radio.com/

                 

                From: sdr-radio-com@yahoogroups.com [mailto:sdr-radio-com@yahoogroups.com] On Behalf Of josephrot

                OH, to be sure, your tests are valid...no argument there, really, HOWEVER...note how many cores you are running and the extreme performance one might need -- can everyone go out and buy 2, 4 or even 8 core hardware, and then run say 6 to 7 instances of SDR ?

                My discussion was more just reminding people to note their performance needs when we may want to run multiple SDR on one machine, but yes, your advisory is welcome news -- multiple SDR can be done.

              • Sandro
                You only need an old I5-2500K CPU (it is a quad core) on a good motherboard and efficient, well written software (the IPP library from Intel helps a lot). 73
                Message 7 of 25 , Aug 6, 2013
                  You only need an "old" I5-2500K CPU (it is a quad core) on a good motherboard and efficient, well written software (the IPP library from Intel helps a lot).

                  73 Sandro


                  --- In sdr-radio-com@yahoogroups.com, "josephrot" <joerotello@...> wrote:
                  >
                  > Leif...Thank you for your comments...
                  >
                  > OH, to be sure, your tests are valid...no argument there, really, HOWEVER...note how many cores you are running and the extreme performance one might need -- can everyone go out and buy 2, 4 or even 8 core hardware, and then run say 6 to 7 instances of SDR ?
                  >
                  > My discussion was more just reminding people to note their performance needs when we may want to run multiple SDR on one machine, but yes, your advisory is welcome news -- multiple SDR can be done.
                  >
                  > Joe
                  >
                  > --- In sdr-radio-com@yahoogroups.com, Leif Asbrink <leif@> wrote:
                  > >
                  > > Hi Joe and all,
                  > >
                  > > I found some statements below a bit provocative. I think modern
                  > > computers have all the processing power we need.
                  > >
                  > > I have tested 7 RTL-SDR dongles running simultaneously with 7 instances
                  > > of Linrad at the maximum sampling rate of 2.4 MHz on a 8-core machine.
                  > > It is not the latest generation, but the CPU load is 43% only and there
                  > > is a lot of resources available for decoders of various kinds. Operation
                  > > seems totally stable and I have no reason to believe there would be any problem
                  > > to run a couple of dozen of decoders.
                  > >
                  > > For unknown reasons Windows is not as efficient. Running with the same
                  > > parameters under Windows 7 causes a CPU load of 90%.
                  > > Operation is stable, but there is not much resources left for decoders.
                  > >
                  > > Note that Linrad runs a full FFT with waterfall for each one of the dongles
                  > > and computes a narrowband signal for a selected frequency in each case.
                  > >
                  > > Surely a single core would not do the work, but if the sampling rate
                  > > were set to 240 kHz a Pentium4 would manage many many dongles:-)
                  > >
                  > > This URL:
                  > > http://www.sm5bsz.com/lir/sdrcmp/bdr/bdr-1.htm
                  > > shows how 5 different hardware are compared in a single computer
                  > > with 5 instances of Linrad running at sampling rates of 96 or 100 kHz.
                  > > The CPU load is 4%!!!
                  > >
                  > > 73
                  > >
                  > > Leif / SM5BSZ
                  >
                • Simon GD4ELI/HB9DRV
                  FWIW I m waiting for the Haswell i5, then I ll buy an Intel NUC box (when they support NUC) for running my software as a station computer (not for development
                  Message 8 of 25 , Aug 6, 2013

                    FWIW I’m waiting for the Haswell i5, then I’ll buy an Intel NUC box (when they support NUC) for running my software as a station computer (not for development use).

                     

                    Nice features of Haswell are less power required and the AVX2 instructions which IPP 7.1 already support.

                     

                    Intel IPP is a big must these days – it’s crazy to write CPU intensive software and not use IPP.

                     

                    Simon GD4ELI/HB9DRV

                    http://v2.sdr-radio.com/

                     

                    From: sdr-radio-com@yahoogroups.com [mailto:sdr-radio-com@yahoogroups.com] On Behalf Of Sandro

                    You only need an "old" I5-2500K CPU (it is a quad core) on a good motherboard and efficient, well written software (the IPP library from Intel helps a lot).

                  • josephrot
                    FYI, wikipedia.org seems to cover or discuss BladeRF in more than one of their entries: http://en.wikipedia.org/wiki/Field_Programmable_RF Also is listed on
                    Message 9 of 25 , Aug 6, 2013
                      FYI, wikipedia.org seems to cover or discuss BladeRF in more than one of their entries:

                      http://en.wikipedia.org/wiki/Field_Programmable_RF

                      Also is listed on the chart at:

                      http://en.wikipedia.org/wiki/List_of_software-defined_radios

                      Interesting info.

                      Joe



                      --- In sdr-radio-com@yahoogroups.com, "Simon GD4ELI/HB9DRV" <simon@...> wrote:
                      >
                      > Leif,
                      >
                      > What do you think about the hackRF & bladeRF? I'm looking forward to the
                      > bladeRF as a possible VHF / UHF / SHF satellite receiver.
                      >
                      > Simon GD4ELI/HB9DRV
                      >
                      > http://v2.sdr-radio.com/
                    • drew231955
                      Hi All I have a Haswell i5 running. Nice chip, lower power and better graphics. One of the biggest problems is the size of ATX cases, even the small ones are
                      Message 10 of 25 , Aug 6, 2013
                        Hi All

                        I have a Haswell i5 running. Nice chip, lower power and better graphics.

                        One of the biggest problems is the size of ATX cases, even the small ones are big.

                        I built an ITX system (150x150 mm motherboard) running Windows 8; much better as the case is about the size of a transceiver. The lower power helps as there is not a lot of cooling and the power supply is small, 150 W.

                        A laptop is a solution, but I like a bigger screen and more options for input/output, especially USB ports.

                        Regards Drew VK4ZXI

                        --- In sdr-radio-com@yahoogroups.com, "Simon GD4ELI/HB9DRV" <simon@...> wrote:
                        >
                        > FWIW I'm waiting for the Haswell i5, then I'll buy an Intel NUC box (when
                        > they support NUC) for running my software as a station computer (not for
                        > development use).
                        >
                        >
                        >
                        > Nice features of Haswell are less power required and the AVX2 instructions
                        > which IPP 7.1 already support.
                        >
                        >
                        >
                        > Intel IPP is a big must these days - it's crazy to write CPU intensive
                        > software and not use IPP.
                        >
                        >
                        >
                        > Simon GD4ELI/HB9DRV
                        >
                        > http://v2.sdr-radio.com/
                        >
                        >
                        >
                        > From: sdr-radio-com@yahoogroups.com [mailto:sdr-radio-com@yahoogroups.com]
                        > On Behalf Of Sandro
                        >
                        > You only need an "old" I5-2500K CPU (it is a quad core) on a good
                        > motherboard and efficient, well written software (the IPP library from Intel
                        > helps a lot).
                        >
                      • Simon GD4ELI/HB9DRV
                        Yes, The Haswell graphics are excellent, something I forgot. These mini-itx systems are very good, www.mini-itx.com . I have an ATOM mini-itx system. Simon
                        Message 11 of 25 , Aug 6, 2013

                          Yes,

                           

                          The Haswell graphics are excellent, something I forgot. These mini-itx systems are very good, www.mini-itx.com . I have an ATOM mini-itx system.

                          Simon GD4ELI/HB9DRV

                          http://v2.sdr-radio.com/

                           

                          From: sdr-radio-com@yahoogroups.com [mailto:sdr-radio-com@yahoogroups.com] On Behalf Of drew231955
                          Sent: 06 August 2013 14:06
                          To: sdr-radio-com@yahoogroups.com
                          Subject: [sdr-radio-com] Re: Multiple SDRs at once

                           

                           

                          Hi All

                          I have a Haswell i5 running. Nice chip, lower power and better graphics.

                          One of the biggest problems is the size of ATX cases, even the small ones are big.

                          I built an ITX system (150x150 mm motherboard) running Windows 8; much better as the case is about the size of a transceiver. The lower power helps as there is not a lot of cooling and the power supply is small, 150 W.

                          A laptop is a solution, but I like a bigger screen and more options for input/output, especially USB ports.

                          Regards Drew VK4ZXI

                          --- In sdr-radio-com@yahoogroups.com, "Simon GD4ELI/HB9DRV" <simon@...> wrote:
                          >
                          > FWIW I'm waiting for the Haswell i5, then I'll buy an Intel NUC box (when
                          > they support NUC) for running my software as a station computer (not for
                          > development use).
                          >
                          >
                          >
                          > Nice features of Haswell are less power required and the AVX2 instructions
                          > which IPP 7.1 already support.
                          >
                          >
                          >
                          > Intel IPP is a big must these days - it's crazy to write CPU intensive
                          > software and not use IPP.
                          >
                          >
                          >
                          > Simon GD4ELI/HB9DRV
                          >
                          > http://v2.sdr-radio.com/
                          >
                          >
                          >
                          > From: sdr-radio-com@yahoogroups.com [mailto:sdr-radio-com@yahoogroups.com]
                          > On Behalf Of Sandro
                          >
                          > You only need an "old" I5-2500K CPU (it is a quad core) on a good
                          > motherboard and efficient, well written software (the IPP library from Intel
                          > helps a lot).
                          >

                        • josephrot
                          Drew, not wanting to go off-topic, but SDR related this is or will be... Presuming you are using ITX or micro-ITX or part component, how much CPU dedicated RAM
                          Message 12 of 25 , Aug 6, 2013
                            Drew, not wanting to go off-topic, but SDR related this is or will be...

                            Presuming you are using ITX or micro-ITX or part component, how much CPU dedicated RAM does your board have room for now, or what's the max ?

                            Secondary question, can the board "take" Win 7, preferably 64bit, as well as Win 8/8.1 ?

                            Just a wondering thought for a possible SDR project that appears capable to include SDR-Radio...Simon might even be able to get royalties for a job well-done.

                            Joe

                            --- In sdr-radio-com@yahoogroups.com, "drew231955" <drew_wollin@...> wrote:
                            >
                            > Hi All
                            >
                            > I have a Haswell i5 running. Nice chip, lower power and better graphics.
                            >
                            > One of the biggest problems is the size of ATX cases, even the small ones are big.
                            >
                            > I built an ITX system (150x150 mm motherboard) running Windows 8; much better as the case is about the size of a transceiver. The lower power helps as there is not a lot of cooling and the power supply is small, 150 W.
                            >
                            > A laptop is a solution, but I like a bigger screen and more options for input/output, especially USB ports.
                            >
                            > Regards Drew VK4ZXI
                          • Simon GD4ELI/HB9DRV
                            What - money? Free money? Where do I sign up? Simon GD4ELI/HB9DRV http://v2.sdr-radio.com/ From: sdr-radio-com@yahoogroups.com
                            Message 13 of 25 , Aug 6, 2013

                              What – money? Free money? Where do I sign up?

                               

                              Simon GD4ELI/HB9DRV

                              http://v2.sdr-radio.com/

                               

                              From: sdr-radio-com@yahoogroups.com [mailto:sdr-radio-com@yahoogroups.com] On Behalf Of josephrot

                              Simon might even be able to get royalties for a job well-done.

                            • Leif Asbrink
                              Hello Simon, ... I do not knbow anything about these hardware. When trying to get info from the Internet I only get sales-talk with no info about the actual
                              Message 14 of 25 , Aug 6, 2013
                                Hello Simon,

                                > What do you think about the hackRF & bladeRF? I'm looking forward to the
                                > bladeRF as a possible VHF / UHF / SHF satellite receiver.

                                I do not knbow anything about these hardware.

                                When trying to get info from the Internet I only
                                get sales-talk with no info about the actual hardware.
                                Uninteresting:-(

                                The units may well be good SDR hardware, but I am too lazy
                                to spend a lot of time trying to locate information about
                                the hardware so I just do not think anything at all about
                                hackRF & bladeRF.

                                73

                                Leif
                              • Calder Latham
                                Leif....check this out...follow the links at the bottom...Cal http://www.kickstarter.com/projects/mossmann/hackrf-an-open-source-sdr-platform W1HHO from the
                                Message 15 of 25 , Aug 6, 2013
                                      Leif....check this out...follow the links at the bottom...Cal
                                   
                                   
                                   

                                  W1HHO from the woods of Maine.

                                   
                                  Sent: Tuesday, August 6, 2013 5:14 PM
                                  Subject: Re: [sdr-radio-com] Re: Multiple SDRs at once
                                   
                                   

                                  Hello Simon,

                                  > What do you think about the hackRF & bladeRF?
                                  I'm looking forward to the
                                  > bladeRF as a possible VHF / UHF / SHF
                                  satellite receiver.

                                  I do not knbow anything about these hardware.

                                  When trying to get info from the Internet I only
                                  get sales-talk with no info about the actual hardware.
                                  Uninteresting:-(

                                  The units may well be good SDR hardware, but I am too lazy
                                  to spend a lot of time trying to locate information about
                                  the hardware so I just do not think anything at all about
                                  hackRF & bladeRF.

                                  73

                                  Leif

                                • drew231955
                                  Hi Joe. It is a mini-ITX (170x170 mm). It uses standard PC parts, including 2 memory slots for up to 16 GB, and a PCI express slot for graphics or whatever
                                  Message 16 of 25 , Aug 6, 2013
                                    Hi Joe.

                                    It is a mini-ITX (170x170 mm). It uses standard PC parts, including 2 memory slots for up to 16 GB, and a PCI express slot for graphics or whatever (although the card can only be one card wide and half height in most mini ITX cases). Card back space could be used for more USB sockets for multiple SDRs. See specs of ASRock H87M-ITX A$125

                                    The dedicated graphics use up to 1792 MB.

                                    I run 8 GB of PC DDR3 memory (2x 4GB) and Windows 8 x64 on most of my systems.

                                    The thing that makes the biggest difference in any system is using a solid state drive (SSD). A 60GB is sufficient for most radio-dedicated systems, but bigger one are faster and much cheaper than they used to be. SSDs make laptops fly, booting in less than 20 seconds.

                                    ITX cases are small, which limits what you can put in them and how much cooling they have. They usually have room for more than one 2.5 SSD or HDD. I use an Antec ISK300-150 -Mini-ITX Desktop case A$99. It uses a slim-line DVD (like a laptop- standard SATA connector, small 5V power connector- took some anguish to find that out). The power supplies vary with small low power internal or external supplies, but some have bigger power supplies. But ITX cases are a good size for radio computers. The smallest mini-ATX cases are still huge.

                                    Regards Drew

                                    --- In sdr-radio-com@yahoogroups.com, "josephrot" <joerotello@...> wrote:
                                    >
                                    > Drew, not wanting to go off-topic, but SDR related this is or will be...
                                    >
                                    > Presuming you are using ITX or micro-ITX or part component, how much CPU dedicated RAM does your board have room for now, or what's the max ?
                                    >
                                    > Secondary question, can the board "take" Win 7, preferably 64bit, as well as Win 8/8.1 ?
                                    >
                                    > Just a wondering thought for a possible SDR project that appears capable to include SDR-Radio...Simon might even be able to get royalties for a job well-done.
                                    >
                                    > Joe
                                    >
                                    > --- In sdr-radio-com@yahoogroups.com, "drew231955" <drew_wollin@> wrote:
                                    > >
                                    > > Hi All
                                    > >
                                    > > I have a Haswell i5 running. Nice chip, lower power and better graphics.
                                    > >
                                    > > One of the biggest problems is the size of ATX cases, even the small ones are big.
                                    > >
                                    > > I built an ITX system (150x150 mm motherboard) running Windows 8; much better as the case is about the size of a transceiver. The lower power helps as there is not a lot of cooling and the power supply is small, 150 W.
                                    > >
                                    > > A laptop is a solution, but I like a bigger screen and more options for input/output, especially USB ports.
                                    > >
                                    > > Regards Drew VK4ZXI
                                    >
                                  • Leif Asbrink
                                    On Tue, 6 Aug 2013 18:15:38 -0400 ... It seems like hack-rf uses a max2837 as the IF. Some of the links do not work on my Linux and I do not have the patience
                                    Message 17 of 25 , Aug 6, 2013
                                      On Tue, 6 Aug 2013 18:15:38 -0400
                                      "Calder Latham" <calder_b@...> wrote:

                                      > Leif....check this out...follow the links at the bottom...Cal
                                      >
                                      >
                                      > http://www.kickstarter.com/projects/mossmann/hackrf-an-open-source-sdr-platform
                                      >
                                      >
                                      > W1HHO from the woods of Maine.
                                      >
                                      It seems like hack-rf uses a max2837 as the IF.
                                      Some of the links do not work on my Linux and I do not
                                      have the patience to really understand what this is.

                                      20 MHz bandwidth on USB 2.0 does not seem a serious claim
                                      to me. Once there is a stable product with whatever performance
                                      I will be interested to include support in Linrad for the Rx side.
                                      Only if performance meets some minimum standards I would be interested
                                      in adding support for the Tx as well. From what I can guess with
                                      what I know now, the tx will not have acceptable noise sidebands
                                      for usage on ham bands so I do not think this hardware will
                                      become an alternative for Tx in Linrad.

                                      For receivers however I have no limit on what is acceptable
                                      performance. Hackrf is a wide coverage rx and I would love to
                                      support it although it is unclear to me how and when stable
                                      hardware will become available.

                                      This thread has been about multiple SDRs and hardware. I just
                                      added the info about running 6 rtl-sdr donmgles simultaneously
                                      on a Pentium 4 single core machine. No problem:-)
                                      http://www.sm5bsz.com/linuxdsp/hware/rtlsdr/rtlsdr.htm

                                      73

                                      Leif / SM5BSZ
                                    • josephrot
                                      Drew, many sincere thanks for the details, and they are both interesting and indicative of one very well-made if not well-designed computer on a card , if I
                                      Message 18 of 25 , Aug 6, 2013
                                        Drew, many sincere thanks for the details, and they are both interesting and indicative of one very well-made if not well-designed "computer on a card", if I read your descriptions right.

                                        The reasoning for asking for the details...and a price point too -- will become evident as I am able to discuss a few ideas with Simon hopefully in the near future.

                                        I humbly agree on SSD and they are already slowly taking over from the rotating HDD's of now yesteryear and today. Hopefully, the SSD's quality and their reaching the 50,000 hours MTBF rating will sooner than later mean the death-knell for rotating drives.

                                        Thank you again for taking the time and effort to really enrich our understanding of the product.

                                        Joe / Knoxville, TN
                                        Slype: joerotello

                                        --- In sdr-radio-com@yahoogroups.com, "drew231955" <drew_wollin@...> wrote:
                                        >
                                        > Hi Joe.
                                        >
                                        > It is a mini-ITX (170x170 mm). It uses standard PC parts, including 2 memory slots for up to 16 GB, and a PCI express slot for graphics or whatever (although the card can only be one card wide and half height in most mini ITX cases). Card back space could be used for more USB sockets for multiple SDRs. See specs of ASRock H87M-ITX A$125
                                        >
                                        > << snipped for reply brevity >>
                                      • claninwood
                                        Simon, regarding your writing of a new PDF. If it helps I am willing to offer to proof read. Chris G7OGX
                                        Message 19 of 25 , Aug 7, 2013
                                          Simon, regarding your writing of a new PDF. If it helps I am willing to offer to proof read. Chris G7OGX
                                        Your message has been successfully submitted and would be delivered to recipients shortly.