Parallel streaming
Designed around multiple samples per clock cycle for applications that require higher FFT throughput.
FPGA signal-processing IP
A configurable feed-forward FFT core under development for high-throughput FPGA applications, extending the published SFF architecture to parallel data paths.
Product development in progress
Architecture
Designed around multiple samples per clock cycle for applications that require higher FFT throughput.
Based on the SFF family of pipeline FFT architectures, with deterministic streaming data flow.
Development targets efficient use of FPGA arithmetic and short-delay storage resources.
The commercial implementation is being developed as a reusable, parameterized FPGA core.
Final supported configurations and implementation results will be published as development progresses.
Technical foundation
C. Ingemarsson and O. Gustafsson, “SFF—The Single-Stream FPGA-Optimized Feedforward FFT Hardware Architecture,” Journal of Signal Processing Systems, vol. 90, pp. 1583–1592, 2018.
The paper introduces the single-stream FPGA-optimized feedforward (SFF) FFT architecture and its FPGA-oriented use of short shift-register storage.
A technical publication describing the parallel SFF extension is planned. A free preprint link will be added here when available.
Contact
Technical and commercial enquiries are welcome while the IP core is in development.
carl@rejmyresp.comRejmyre SP AB · Sweden