I need to compute the FFT of 300,000 elements of fixed point type (FFT length of 150,000). On my current FPGA (Kintex‑7 325T FPGA), when using the FFT express function I have a grayed out limit of 65,536, and the largest fft length actually available to me is 8,192. I know there are ways to work around this by breaking up my data into smaller DFT's and then recombining them, or programming my own implementation (which would surely be less efficient than LabVIEW's FFT function?) But I assume the issue is lack of resources, and I believe either approach would run into the same sort of trouble. Is there some way to overcome this on my current FPGA, and if not, then what sort of specs should I be looking for when purchasing a new one?
