The Pipelined Streaming I/O solution pipelines several Radix-2 butterfly processing engines to offer continuous data processing. Each processing engine has its own memory banks to store the input and intermediate data ( This Figure ). The core has the ability to simultaneously perform transform calculations on the current frame of data, load input data for the next frame of data, and unload the results of the previous frame of data. You can continuously stream in data and, after the calculation latency, can continuously unload the results. If preferred, this design can also calculate one frame by itself or frames with gaps in between.
IMPORTANT: Continually streaming data does not imply that AXI4-Stream waitstates from the FFT core can be ignored. There are situations where the FFT core might have to insert waitstates to pause the incoming sample data.
In the scaled fixed-point mode, the data is scaled after every pair of Radix-2 stages. The block floating-point mode might use significantly more resources than the scaled mode, as it must maintain extra bits of precision to allow dynamic scaling without impacting performance. Therefore, if the input data is well understood and is unlikely to exhibit large amplitude fluctuation, using scaled arithmetic (with a suitable scaling schedule to avoid overflow in the known worst case) is sufficient, and resources might be saved.
The input data is presented in natural order. The unloaded output data can either be in bit reversed order or in natural order. When natural order output data is selected, additional memory resource is utilized.
This architecture covers point sizes from 8 to 65536. You have the flexibility to select the number of stages to use block RAM for data and phase factor storage. The remaining stages use distributed memory.
