CRIHAN  - Mars 2001
Outils d’optimisation - Page 39
4- F. application : ADI
•résultat de perfex -a -y -x
•                                                                           Typical      Minimum      Maximum
•   Event Counter Name                                      Counter Value   Time (sec)   Time (sec)   Time (sec)
•===============================================================================================================
• 0 Cycles..................................................    711044336     3.646381     3.646381     3.646381
•16 Cycles..................................................    711044336     3.646381     3.646381     3.646381
•23 TLB misses..............................................      5604048     1.956819     1.956819     1.956819
•26 Secondary data cache misses.............................      1061616     0.411036     0.268725     0.457312
• 7 Quadwords written back from scache......................      6004864     0.197083     0.130259     0.197083
•25 Primary data cache misses...............................      4235360     0.195695     0.061250     0.195695
•21 Graduated floating point instructions...................     29959296     0.153637     0.076819     7.989146
•22 Quadwords written back from primary data cache..........      6328832     0.124954     0.101910     0.144427
• 9 Primary instruction cache misses........................        42816     0.003957     0.001236     0.003957
•10 Secondary instruction cache misses......................          896     0.000347     0.000227     0.000386
•Statistics
•=====================================================================================
•Graduated instructions/cycle............................................     0.456588
•Graduated floating point instructions/cycle.............................     0.042134
•Graduated loads & stores/cycle..........................................     0.155101
•Graduated loads & stores/floating point instruction.....................     3.681107
•L1 Cache Line Reuse.....................................................    25.038722
•L2 Cache Line Reuse.....................................................     2.989540
•L1 Data Cache Hit Rate..................................................     0.961596
•L2 Data Cache Hit Rate..................................................     0.749345
•Time accessing memory/Total time........................................     0.858140
•L1--L2 bandwidth used (MB/s, average per process).......................    64.939132
•Memory bandwidth used (MB/s, average per process).......................    63.615036
•MFLOPS (average per process)............................................     8.216172