This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
在TI提供的一份测试报告中,有如下描述:
Performance Benchmarks
The full benefit of optimized code can be achieved if the data is not partitioned into small buffers and if at least ping, pong, pang, and peng buffers are in internal memory. The performance of 777 cycles per edge point (or 39 cycles per pixel, assuming 5% of pixels are edge points) has been achieved, with input and output data in external memory and ping, pong, pang, and peng buffers in internal memory. The number of edge points in this measurement was 3840 (5% of 320x240 image), while the size of the Hough Space in this measurement was 267x267.
但是我没有看到full benefit of optimized的结果是什么,请问使用internal memory和external memory的区别有多大?如何将数组分别开辟在internal memory和external memory? 仿真的时候,是否也有能测试internal memory和external memory的区别? 多谢了!
Chris Meng,谢谢!
我使用的是tda3x芯片中的dsp,是C66的DSP。请问内部memory是集成在芯片中,外部memory是芯片通过针脚访问的外部memory吗?
ccs 仿真的时候,是不是都是使用的内部memory?
如果是这样,那么在使用ccs仿真的时候,是否就无法测试出内部memory和外部memory的区别?
Zhang Hui,
你说的仿真是使用仿真器链接板子,并且把程序下载到板上么?如果是,这是可以访问外部memory,例如DDR的。代码具体放在哪里,由你下载的程序决定。