cactuBSSN is a SPEC CPU(R) benchmark described here and written in C, C++ and Fortran. The workload runs on all logical cores.
Topdown profile shows this is a backend-bound workload.
AMD metrics confirm the benchmark is memory bound. Only ~55 floating point instructions per 1000. Not many branches. There is a large percentage of L2 access including ~10% L2 misses.
elapsed 609.922
on_cpu 0.981 # 15.69 / 16 cores
utime 9531.490
stime 37.821
nvcsw 15443 # 11.89%
nivcsw 114431 # 88.11%
inblock 0 # 0.00/sec
onblock 302096 # 495.30/sec
cpu-clock 9571161124089 # 9571.161 seconds
task-clock 9571308947593 # 9571.309 seconds
page faults 10865926 # 1135.260/sec
context switches 129126 # 13.491/sec
cpu migrations 250 # 0.026/sec
major page faults 1384 # 0.145/sec
minor page faults 10864542 # 1135.116/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 458027626592 # 48.787 branches per 1000 inst
branch misses 2930723831 # 0.64% branch miss
conditional 313602888445 # 33.403 conditional branches per 1000 inst
indirect 35057805161 # 3.734 indirect branches per 1000 inst
cpu-cycles 41339157355858 # 4.21 GHz
instructions 9387529160935 # 0.23 IPC low
slots 82669018078590 #
retiring 3387165858961 # 4.1% ( 4.3%) low
-- ucode 385306245 # 0.0%
-- fastpath 3386780552716 # 4.1%
frontend 3820302772231 # 4.6% ( 4.8%) low
-- latency 2938260448878 # 3.6%
-- bandwidth 882042323353 # 1.1%
backend 72418746717494 # 87.6% (90.9%) high
-- cpu 7511969518602 # 9.1%
-- memory 64906777198892 # 78.5%
speculation 54605694491 # 0.1% ( 0.1%) low
-- branch mispredict 33703731722 # 0.0%
-- pipeline restart 20901962769 # 0.0%
smt-contention 2988168305306 # 3.6% ( 0.0%)
cpu-cycles 41172189685093 # 4.21 GHz
instructions 9393665948727 # 0.23 IPC low
instructions 3125620100025 # 476.467 l2 access per 1000 inst
l2 hit from l1 1102457333145 # 10.76% l2 miss
l2 miss from l1 96542777657 #
l2 hit from l2 pf 323167259778 #
l3 hit from l2 pf 24513021077 #
l3 miss from l2 pf 39118477073 #
instructions 3128526163741 # 55.472 float per 1000 inst
float 512 308 # 0.000 AVX-512 per 1000 inst
float 256 1236272 # 0.000 AVX-256 per 1000 inst
float 128 173544139570 # 55.472 AVX-128 per 1000 inst
float MMX 0 # 0.000 MMX per 1000 inst
float scalar 1 # 0.000 scalar per 1000 inst
instructions 9383775361054 #
opcache 1267171993600 # 135.039 opcache per 1000 inst
opcache miss 604865222155 # 47.7% opcache miss rate
l1 dTLB miss 534977958244 # 57.011 L1 dTLB per 1000 inst
l2 dTLB miss 7159514212 # 0.763 L2 dTLB per 1000 inst
instructions 9383603598466 #
icache 648567980402 # 69.117 icache per 1000 inst
icache miss 271389699675 # 41.8% icache miss rate
l1 iTLB miss 569831957 # 0.061 L1 iTLB per 1000 inst
l2 iTLB miss 0 # 0.000 L2 iTLB per 1000 inst
tlb flush 106503 # 0.000 TLB flush per 1000 inst
Process overview shows cactusBSSN_r_ba as the primary process.
775 processes
48 cactusBSSN_r_ba 9598.22 27.75
165 specperl 25.72 3.22
41 specinvoke 0.01 0.00
1 clang++ 0.01 0.00
1 flang 0.01 0.00
1 lsb_release 0.01 0.00
1 clang 0.00 0.01
270 sh 0.00 0.00
54 specrxp 0.00 0.00
48 bash 0.00 0.00
21 grep 0.00 0.00
20 cat 0.00 0.00
12 uniq 0.00 0.00
11 sort 0.00 0.00
10 expand 0.00 0.00
10 ps 0.00 0.00
6 pwd 0.00 0.00
5 basename 0.00 0.00
5 specmake 0.00 0.00
5 systemctl 0.00 0.00
4 specpp 0.00 0.00
4 uname 0.00 0.00
3 dirname 0.00 0.00
3 dmidecode 0.00 0.00
3 lscpu 0.00 0.00
2 df 0.00 0.00
2 dpkg 0.00 0.00
2 rm 0.00 0.00
2 runcpu 0.00 0.00
2 specsha512sum 0.00 0.00
2 specxz 0.00 0.00
2 who 0.00 0.00
1 cpupower 0.00 0.00
1 head 0.00 0.00
1 logname 0.00 0.00
1 ls 0.00 0.00
1 numactl 0.00 0.00
1 sysctl 0.00 0.00
1 w 0.00 0.00
1 wc 0.00 0.00
1 which 0.00 0.00
0 processes running
53 maximum processes
Computation blocks are simple with the spec harness invoking one copy on each logical core.
371157) specinvoke cpu=10 start=3.69 finish=206.29
371159) sh cpu=5 start=3.69 finish=205.12
371166) bash cpu=0 start=3.69 finish=205.12
371191) cactusBSSN_r_ba cpu=0 start=3.69 finish=205.05
371160) sh cpu=9 start=3.69 finish=204.08
371168) bash cpu=1 start=3.69 finish=204.08
371197) cactusBSSN_r_ba cpu=1 start=3.69 finish=203.99
371161) sh cpu=2 start=3.69 finish=203.74
371169) bash cpu=2 start=3.69 finish=203.74
371192) cactusBSSN_r_ba cpu=2 start=3.69 finish=203.61
371162) sh cpu=7 start=3.69 finish=203.55
371174) bash cpu=3 start=3.69 finish=203.55
371193) cactusBSSN_r_ba cpu=3 start=3.69 finish=203.45
371163) sh cpu=13 start=3.69 finish=206.29
371175) bash cpu=4 start=3.69 finish=206.29
371196) cactusBSSN_r_ba cpu=4 start=3.69 finish=206.23
371164) sh cpu=9 start=3.69 finish=204.12
371178) bash cpu=5 start=3.69 finish=204.12
371194) cactusBSSN_r_ba cpu=5 start=3.69 finish=204.02
371165) sh cpu=10 start=3.69 finish=204.49
371179) bash cpu=6 start=3.69 finish=204.49
371195) cactusBSSN_r_ba cpu=6 start=3.69 finish=204.40
371167) sh cpu=7 start=3.69 finish=202.92
371176) bash cpu=7 start=3.69 finish=202.92
371199) cactusBSSN_r_ba cpu=7 start=3.69 finish=202.80
371170) sh cpu=9 start=3.69 finish=205.12
371181) bash cpu=8 start=3.69 finish=205.12
371198) cactusBSSN_r_ba cpu=8 start=3.69 finish=205.05
371171) sh cpu=10 start=3.69 finish=204.07
371182) bash cpu=9 start=3.69 finish=204.07
371201) cactusBSSN_r_ba cpu=9 start=3.69 finish=203.97
371172) sh cpu=3 start=3.69 finish=203.74
371185) bash cpu=10 start=3.69 finish=203.74
371202) cactusBSSN_r_ba cpu=10 start=3.69 finish=203.61
371173) sh cpu=11 start=3.69 finish=203.56
371186) bash cpu=11 start=3.69 finish=203.55
371200) cactusBSSN_r_ba cpu=11 start=3.69 finish=203.45
371177) sh cpu=10 start=3.69 finish=206.29
371187) bash cpu=12 start=3.69 finish=206.29
371203) cactusBSSN_r_ba cpu=12 start=3.69 finish=206.23
371180) sh cpu=10 start=3.69 finish=204.12
371188) bash cpu=13 start=3.69 finish=204.12
371204) cactusBSSN_r_ba cpu=13 start=3.70 finish=204.02
371183) sh cpu=1 start=3.69 finish=204.48
371189) bash cpu=14 start=3.69 finish=204.48
371205) cactusBSSN_r_ba cpu=14 start=3.70 finish=204.40
371184) sh cpu=14 start=3.69 finish=204.69
371190) bash cpu=15 start=3.69 finish=204.69
371206) cactusBSSN_r_ba cpu=15 start=3.70 finish=204.62