A Kernel Interleaved Scheduling Method for Streaming Applications on Soft-core Vector Processors
Published in 2011 IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA), pp. 278–285, 2011
This paper addresses the challenge of efficiently scheduling streaming applications on soft-core vector processors implemented on FPGA platforms. The authors introduce the Kernel Interleaved Scheduling (KIS) method, which aims to maximize resource utilization and enhance processing throughput by interleaving kernels at runtime.
Key contributions include:
Kernel Interleaved Scheduling (KIS): A method for interleaving multiple computational kernels to improve resource utilization and parallelism.
FPGA Soft-Core Vector Processors: Application of the method to FPGA-based soft-core vector processors, demonstrating flexibility and scalability.
Performance Evaluation: Experimental validation showing significant improvements in throughput and efficiency for streaming applications compared to traditional scheduling approaches.
Recommended citation: Zheng, C., McAllister, J., & Wu, Y. (2011). A kernel interleaved scheduling method for streaming applications on soft-core vector processors. In *2011 IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA)* (pp. 278–285). IEEE. https://doi.org/10.1109/RTCSA.2011.6006617
Download Paper | Download Slides | Download Bibtex