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.

TMS320C6670: the principles of caches

Part Number: TMS320C6670

Hi all,

I have two questions about cache principle.

(1) from the document, DSP Cache User Guide, http://www.ti.com/lit/ug/sprugy8/sprugy8.pdf

If i put all of my program/data in L2 SRAM

page:1-13 shows L2 data/program needs to be fetched through L1, then CPU execute the code based on the program address in L2.

However, my colleague says he disables caches in L1 data(max L1 data SRAM) and disable SRAM in L1 program(max L1 program caches), so the situation should be CPU directly executes code in L2 and it doesn't matter if L1 has those data/program.

is this statement true?

(2) why we need two different layers, L1 and L2? Does CPU have different speed to fetch data/program in L1 and L2?

If they are the same, why not CPU directly access program and data in L2?

Thank you for response.