优化TMS320F2812内部FLASH运算效率的研究  

Research on Improving the Execution Efficiency of TMS320F2812 Internal FLASH

作　　者：樊燕红[1] 李晔[1] 姜竞赛[1] 张鹏[1] 马晓凤[1] 郝秋赟[1] 

机构地区：[1]山东省计算中心山东省计算机网络重点实验室,山东济南250101

出　　处：《电声技术》2014年第8期62-64,共3页Audio Engineering

基　　金：山东省自然科学基金项目(ZR2010FW008);国家国际科技合作专项项目(2012DFR10500)

摘　　要：在实际应用中,因为TMS320F2812内部RAM的空间限制,有些复杂的程序算法需要运行在内部FLASH中。提高FLASH的运算效率,对于算法的实现具有重要意义。FLASH的运算效率受三个因素的影响:FLASH流水线、取值等待时间和程序执行代码。针对这三个影响因素提出了三种解决方法,并应用到实际算法中。测试结果显示,在低速率语音编解码算法运行时,FLASH的运算效率达到了77.35 MI/s。In practical applications, because TMS320F2812 internal RAM space constraints, some complex algorithms are required to run the program in the internal FLASH. Improving execution efficiency of FLASH has important significance to the algorithm. Execution efficiency of FLASH is affected by three factors of FLASH pipeline, fetch access wait states and program execution code. According to the three factors, three solutions are given. The solutions are applied to practical algorithms. Test results show that in the low -rate speech codec running algorithm, FLASH operational efficiency reached 77.35 MI/s.

关 键 词：FLASH流水线 取值等待时间 执行效率 代码线性度 

分 类 号：TN912.3[电子电信—通信与信息系统]