收录:
摘要:
The Very Long Instruction Word (VLIW) architecture is widely used for DSP. In the hybrid coding process, since the procedures of inverse quantization and inverse scan are frequently called and the methods of optimizing the loops that frequently appears in programs are studied. With TM1300 as an example, the character of the VLIW architecture is introduced. Optimizing these two procedures would improve the speed of whole decoding process. These methods are applied to optimize the inverse scan and inverse quantization in the MPEG4. The instruction cycles required for the optimized algorithm is about 60% of the original one.
关键词:
通讯作者信息:
电子邮件地址: