Programmability in Dictionary-Based Compression

The size of the program code has become a critical design constraint in embedded systems. Large program codes require large memories, which increase the size and cost of the chip. Furthermore, large memories may increase the power consumption due to higher memory I/O bandwidth. Program compression methods have been proposed to improve the code density and hence minimize the area and power consumption of the program memory. Dictionary-based compression is one of the most often used compression methods due to its simplicity. The drawback of the dictionary-based compression is the limited programmability after compression when the dictionary is implemented using ROM or standard cells. In this paper, a method based on extending the dictionary is proposed for a customizable processor architecture with parallel resources to maintain the programmability also after the compression has been made. The obtained results indicate that with the proposed approach, the programmability can be maintained with a small overhead in area and power consumption