•résultats de perfex -a -y -x :
•
Typical
Minimum
Maximum
•Event Counter
Name
Counter Value Time
(sec) Time (sec) Time (sec)
•===============================================================================================================
• 0
Cycles.................................................. 1682480768 8.628107 8.628107 8.628107
•16
Cycles.................................................. 1682480768 8.628107 8.628107 8.628107
•26
Secondary data cache misses............................. 6573952 2.545299 1.664053 2.831856
•23 TLB
misses.............................................. 5787712 2.020950 2.020950 2.020950
• 7 Quadwords written
back from scache...................... 52204928 1.713393 1.132445 1.713393
•25
Primary data cache misses............................... 17305104 0.799585 0.250258 0.799585
•22 Quadwords written
back from primary data cache.......... 31548640 0.622883 0.508014 0.719956
•21 Graduated
floating point instructions................... 30801216 0.157955 0.078977 8.213658
• 9 Primary
instruction cache misses........................ 15040 0.001390 0.000434 0.001390
•10 Secondary
instruction cache misses...................... 544 0.000211 0.000138 0.000234
•Statistics
•=====================================================================================
•Graduated
instructions/cycle............................................ 0.205890
•Graduated floating
point instructions/cycle............................. 0.018307
•Graduated loads
& stores/cycle.......................................... 0.074188
•Graduated loads
& stores/floating point instruction..................... 4.052416
•L1
Cache Line Reuse..................................................... 6.212863
•L2
Cache Line Reuse..................................................... 1.632375
•L1 Data
Cache Hit Rate.................................................. 0.861359
•L2 Data
Cache Hit Rate.................................................. 0.620115
•Time accessing
memory/Total time........................................ 0.696089
•L1--L2 bandwidth
used (MB/s, average per process)....................... 122.685269
•Memory bandwidth
used (MB/s, average per process)....................... 194.335189
•MFLOPS (average per
process)............................................ 3.569870