https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/DFT-verses-FFT-performance-for-non-2-sizes/m-p/988069#M17884 <DIV> <DIV><FONT color="#000000">I am migrating from the MKLFFT calls to the MKL DFT calls using C++. I have a fft size that is 700K long that requires me to use a fft length of 2^20th. With the DFT call in MKL vs the FFT call will I see a 25% speed increase because I can use the 700K sample fft in the DFT instead of the 2^20th for the MKL FFT?</FONT></DIV></DIV> Thu, 23 Jun 2005 02:25:57 GMT Intel_C_Intel 2005-06-23T02:25:57Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/DFT-verses-FFT-performance-for-non-2-sizes/m-p/988069#M17884 <DIV> <DIV><FONT color="#000000">I am migrating from the MKLFFT calls to the MKL DFT calls using C++. I have a fft size that is 700K long that requires me to use a fft length of 2^20th. With the DFT call in MKL vs the FFT call will I see a 25% speed increase because I can use the 700K sample fft in the DFT instead of the 2^20th for the MKL FFT?</FONT></DIV></DIV> Thu, 23 Jun 2005 02:25:57 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/DFT-verses-FFT-performance-for-non-2-sizes/m-p/988069#M17884 Intel_C_Intel 2005-06-23T02:25:57Z