Configuration Compression for Virtex FPGAs
Zhiyuan Li, Scott Hauck · 2001
Although run-time reconfigurable systems have been shown to achieve very high performance, the speedups over traditional microprocessor systems are limited by the cost of configuration of the hardware. Current reconfigurable systems suffer from a significant overhead due to the time it takes to reconfigure their hardware. In order to deal with this overhead, and increase the compute power of reconfigurable systems, it is important to develop hardware and software systems to reduce or eliminate this delay. In this paper, we explore the idea of configuration compression and develop algorithms for reconfigurable systems. These algorithms, targeted to Xilinx Virtex series FPGAs with minimum modification of hardware, can significantly reduce the amount of data needed to transfer during configuration. In this work we have extensively investigated current compression techniques, including Huffman coding, Arithmetic coding and LZ coding. We have also developed different algorithms targeting different hardware structures. Our readback algorithm allows certain frames to be reused as a dictionary. In addition, we have developed frame reordering techniques that better uses the regularities by shuffling the sequence of the configuration. We have also developed a wildcard approach that can be used for true partial reconfiguration. The simulation results demonstrate that a factor of 4 compression ratio can be achieved.