Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am using multicore simulator marss http://marss86.org. I am using 4-core simulation, with shared L2 cache. Each core is run for say, 100M instructions. Since different cores progress at different rates, I collect stats only for 100M instruction of each core. The overall simulation runs till the last-to-finish core has done 100M instructions.

I wanted to calculate MPKI (miss-per-Kilo-instruction) for L2. I was not clear, whether MPKI should be

Sum_Of_L2_Miss_Of_All_Cores/ 400M

or

Sum_Of_L2_Miss_Of_All_Cores/ 100M

or something else. Can you clarify? Thanks.

share|improve this question
up vote 1 down vote accepted

The MPKI is

Sum_Of_L2_Miss_Of_All_Cores/ 400K

(note the K--its MP*K*I not misses per instruction) since 400M instructions were executed. It might help to imagine the workload being run on one core with very fine-grained context switching with no OS overhead and with per thread miss counts maintained.

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .