I am having a hard time getting accurate reading of the cache hits and misses for Brodwell. The output always shows the cache misses to be significantly greater than the cache hits. Is there a way to fix this problem? Is there some setting that needs to be turned on to get this data?
- Parallel Computing
I had an issue in the past with cache misses and the following thread provided me with the right option. I was making the same mistake mentioned in the thread and I applied the solution marked as Best Reply.
Take a look at the thread. It might probably help you: https://software.intel.com/en-us/forums/software-tuning-performance-optimization-platform-monitoring/topic/280192