Efficient Cache and BLOB - View / Skip Profiling Cache

Question

For the cache to work effectively, should the data used be stored linearly correctly?

So instead of dynamically distributing, I put my data in blob using a linear allocator. Is that enough to improve performance? What should I do to improve caching efficiency?

I know this question is specific, but I don’t know how to explain it ...

What programs can help me miss or miss caching?

Answer 1

, AMD CodeAnalyst VerySleepy, , AMD ( Intel, iirc ), , . , , , , , Agner Fog Intel ( )

Answer 2

Linux, Valgrind ( , cachegrind).

Windows, VS2010 (2008) Professional Edition , . Intel VTune Analyzer (). , , 30- .

SO, :

• ( ++)
• C ++

Answer 3

Linux perf mem , ( ), .