Adaptive Parallel Optimization Design for Signal Processing on Multi-core and Many-core Platforms

Based on the characteristics and parallel design research of multi-core and many-core hardware platforms, this paper proposes an adaptive parallel optimization design method for signal processing on multi-core and many-core platforms. By employing a hardware platform architecture-aware approach and a multi-core segmented pipeline design, adaptive parallel optimization is achieved. The method was tested and verified using a two-stage FFT algorithm on the Phytium 2000+ platform. The test results indicate that through multiple iterations of optimization, adaptive high-performance parallel processing on multi-core and many-core platforms can be realized.