I found out that the latest version of the Casper/Astron "CAStron" FFT isn't currently on the main github branch. The plots on my web page have now been updated using the latest version, on the "talon_sync_test" branch, which achieves a higher clock speed of just under 500MHz on the tested FPGA for smaller FFT sizes.
Thanks to Talon Myburgh for helping me to get it working for these tests. Regards, Ross On Thu, Apr 17, 2025, 1:25 PM Ross Martin <ross.mar...@ieee.org> wrote: > Hi Everyone, > > I've expanded my comparison of FFT performance in Ultrascale to include > some new FFTs. > > You can see the plots at https://bxbfft.com/Ultrascale.html. > > The comparison now includes > > Astron FFT from opencores.org > BxBFFT https://bxbfft.com > CAStron FFT https://github.com/talonmyburgh/casper_dspdevel.git > htfft https://github.com/benreynwar/htfft.git > SGen FFT https://acl.inf.ethz.ch/research/hardware/ > Spiral FFThttps://www.spiral.net/hardware/dftgen.html > Xilinx SSR FFT in Vitis > ZipCPU FFT https://github.com/ZipCPU/dblclockfft > > It doesn't include the original CASPER FFT, since I don't know how to > get RTL source code for it. It doesn't include other commercial FFTs > for the same reason. > > If anyone knows of other FFTs that I should include, or other ways that > I should compare FFTs, let me know. > > I also have plots up for FFTs in Xilinx Versal and for FFTs in various > Altera FPGAs, but those plots don't currently include as many FFTs. > I'll update them as I have time. > > Regards, > > Ross > > r...@bitbybitsp.com > > > -- You received this message because you are subscribed to the Google Groups "casper@lists.berkeley.edu" group. To unsubscribe from this group and stop receiving emails from it, send an email to casper+unsubscr...@lists.berkeley.edu. To view this discussion visit https://groups.google.com/a/lists.berkeley.edu/d/msgid/casper/CAG4nf72vo8PGOWW_GhNBYfQ9cLYP3Za1X4%3D3wpjBy%2BV4TGjEHg%40mail.gmail.com.
