Hi Andrew, I didn't find the "tick" for 'Use DSP48s for adders' in the fft's parameters. where is it located?
Chao-Te -- Open WebMail Project (http://openwebmail.org) ---------- Original Message ----------- From: Andrew Martens <[email protected]> To: Mark Wagner <[email protected]> Cc: Chao-Te Li <[email protected]>, [email protected] Sent: Mon, 3 May 2010 10:14:20 +0200 Subject: Re: [casper] casper Digest, Vol 30, Issue 1 > Hi Chao-Te > > To save on DSP48Es in the FFT try the following > > Un-tick the 'Use DSP48s for adders' (uses the adders in DSP48 blocks > to implement the twiddle stage in the FFT). Should save you 4 > DSP48Es per FFT stage. Set the 'Optimization Target' to > 'Multipliers' (reduces the use of multipliers in the complex > multiplication). Should save you 1 DSP48E per FFT stage. > > Good luck > > Andrew > > On 3 May 2010 07:59, Mark Wagner <[email protected]> wrote: > > > Hi Chao-Te, > > > > The 100% utilization of DSP48's seems to be your problem. I would try > > using slices instead. > > In the fft and pfb paramater box, you can do this by checking 'specify > > multiplier use' and then entering in a vector that specifies the multipliers > > you want to use. I think you want 'behavioral,' so if you have a 2^7 fft, > > you would enter in: [2 2 2 2 2 2 2]. > > > > Mark > > > > > > On Sun, May 2, 2010 at 10:12 PM, Chao-Te Li <[email protected] that > > unless you dig deep into the design, you're nota.edu.tw> wrote: > > > >> Hi Mark, > >> > >> Thanks for the help. > >> I copy the utility summary from MATLAB as follows - > >> > >> Device Utilization Summary: > >> > >> Number of BUFGs 3 out of 32 9% > >> Number of DCM_ADVs 1 out of 12 8% > >> Number of DSP48Es 64 out of 64 100% > >> Number of ILOGICs 65 out of 800 8% > >> Number of External IOBs 97 out of 640 15% > >> Number of LOCed IOBs 97 out of 97 100% > >> > >> Number of External IOBMs 1 out of 320 1% > >> Number of LOCed IOBMs 1 out of 1 100% > >> > >> Number of External IOBSs 1 out of 320 1% > >> Number of LOCed IOBSs 1 out of 1 100% > >> > >> Number of OLOGICs 5 out of 800 1% > >> Number of RAMB18X2s 50 out of 148 33% > >> Number of RAMB36_EXPs 14 out of 148 9% > >> Number of Slice Registers 21382 out of 69120 30% > >> Number used as Flip Flops 21382 > >> Number used as Latches 0 > >> Number used as LatchThrus 0 > >> > >> Number of Slice LUTS 21127 out of 69120 30% > >> Number of Slice LUT-Flip Flop pairs 27717 out of 69120 40% > >> > >> ------------------------------------------------------------------ > >> > >> The model does use up all the DSP48Es (and some other devices). > >> But they are at most 100%. So it's ok in terms of resources? > >> > >> Regards, > >> Chao-Te > >> > >> > >> -- > >> Open WebMail Project (http://openwebmail.org) > >> > >> > >> ---------- Original Message ----------- > >> From: Mark Wagner <[email protected]> > >> To: Chao-Te Li <[email protected]> > >> Cc: [email protected] > >> Sent: Sun, 2 May 2010 21:48:48 -0700 > >> Subject: Re: [casper] casper Digest, Vol 30, Issue 1 > >> > >> > Hi Chao-Te, > >> > > >> > To answer you question, the xps executable should be in the EDK > >> directory. > >> > For instance, ours is located here: > >> > > >> > /opt/Xilinx/11.1/EDK/bin/lin64/xps > >> > > >> > I haven't directly used xps on windows, but I think it should be > >> something > >> > like this: > >> > > >> > C:\Xilinx\10.1\EDK\bin\nt\xps.exe > >> > > >> > >From the XPS_ROACH_base directory on linux I run the command: > >> > > >> > $xps -nw system.xmp > >> > > >> > Although, before you try to remove this error, I would check to see > >> > what the utilization report looks like and if you actually have the > >> > resources to build the design. You can find this report in: > >> > > >> > <model_directory>/XPS_ROACH_base/implementation/system_map.mrp > >> > > >> > The example spectrometer was designed for the sx95t and may not fit > >> > on the lx110t (which doesn't have very many dsp48's). > >> > > >> > Mark > >> > > >> > On Sun, May 2, 2010 at 9:12 PM, Chao-Te Li > >> > <[email protected]> wrote: > >> > > >> > > Hi, > >> > > > >> > > I was compiling the wideband spectrometer from the 2009 tutorials. > >> > > http://casper.berkeley.edu/wiki/Tutorials > >> > > > >> > > It was compilted well with ROACH:sx95t. but when I chose > >> > > ROACH:lx110t, I got the following error message and suggestion - > >> > > > >> > > > >> > > xilperl D:/Xilinx/EDK/data/fpga_impl/observe_par.pl -error yes > >> > > implementation/system.par > >> > > Analyzing implementation/system.par > >> > > > >> > > > >> > >> ***************************************************************************** > >> > > *** > >> > > ERROR: 1 constraint not met. > >> > > > >> > > PAR could not meet all timing constraints. A bitstream will not be > >> > > generated. > >> > > > >> > > To disable the PAR timing check: > >> > > > >> > > 1> Disable the "Treat timing closure failure as error" option from the > >> > > Project Options dialog in XPS. > >> > > > >> > > OR > >> > > > >> > > 2> Type following at the XPS prompt: > >> > > XPS% xset enable_par_timing_error 0 > >> > > > >> > > > >> > > -------------------------------------------------------------------- > >> > > > >> > > > >> > > How do I do to bring up the XPS or its prompt? > >> > > Thanks. > >> > > > >> > > > >> > > -- > >> > > Open WebMail Project (http://openwebmail.org) > >> > > > >> > > > >> > > ---------- Original Message ----------- > >> > > From: [email protected] > >> > > To: [email protected] > >> > > Sent: Sat, 01 May 2010 12:55:16 -0700 > >> > > Subject: casper Digest, Vol 30, Issue 1 > >> > > > >> > > > Send casper mailing list submissions to > >> > > > [email protected] > >> > > > > >> > > > To subscribe or unsubscribe via the World Wide Web, visit > >> > > > > >> > > > >> > > https://calmail.berkeley.edu/manage/list/listinfo/[email protected] > >> > > ey.edu > >> > > > > >> > > > or, via email, send a message with subject or body 'help' to > >> > > > [email protected] > >> > > > > >> > > > You can reach the person managing the list at > >> > > > [email protected] > >> > > > > >> > > > When replying, please edit your Subject line so it is more specific > >> > > > than "Re: Contents of casper digest..." > >> > > > > >> > > > Today's Topics: > >> > > > > >> > > > 1. Re: 'The Great Debate: Are We Alone?' (Tonight) (dana whitlow) > >> > > > > >> > > > > >> ---------------------------------------------------------------------- > >> > > > > >> > > > Message: 1 > >> > > > Date: Fri, 30 Apr 2010 17:15:22 -0400 > >> > > > From: dana whitlow <[email protected]> > >> > > > Subject: Re: [casper] 'The Great Debate: Are We Alone?' (Tonight) > >> > > > To: melvyn wright <[email protected]> > >> > > > Cc: [email protected] > >> > > > Message-ID: <[email protected]> > >> > > > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > >> > > > > >> > > > Good idea- we'll just hook that right up. Let's see now- should it > >> > > > be 430 MHz or 2380 MHz? > >> > > > > >> > > > Dana > >> > > > > >> > > > melvyn wright wrote: > >> > > > > On Fri, Apr 30, 2010 at 12:13 PM, David Hawkins > >> <[email protected]> > >> > > wrote: > >> > > > > > >> > > > >>> will the debate be recorded, and available later for humans or > >> for > >> > > LGMs ? > >> > > > >>> > >> > > > >> Shouldn't we be blasting it out into space for the LGMs > >> > > > >> to listen to live? > >> > > > >> > >> > > > >> :) > >> > > > >> > >> > > > > > >> > > > > not according to > >> > > > > > >> http://www.csmonitor.com/USA/Politics/DC-Decoder/2010/0430/Stephen- > >> > > Hawking-alien-warning-Could-we-talk-to-them > >> > > > > > >> > > > > :-( > >> > > > > > >> > > > > M. > >> > > > > > >> > > > > > >> > > > > > >> > > > > > >> > > > > >> > > > End of casper Digest, Vol 30, Issue 1 > >> > > > ************************************* > >> > > ------- End of Original Message ------- > >> > > > >> > > > >> > > > >> ------- End of Original Message ------- > >> > >> > > ------- End of Original Message -------
