Livermore OpenMP test with one workload test. Looks to be mostly single-threaded with short sections of multi-threaded runs.
Topdown profile looks backend bound with the short parallel sections less so.
AMD metrics show an average of only 3.5 cores. This is floating point code with few branch misses. Frontend stalls are low and backend stalls are high.
elapsed 546.949
on_cpu 0.218 # 3.49 / 16 cores
utime 1904.013
stime 2.173
nvcsw 16236 # 51.16%
nivcsw 15499 # 48.84%
inblock 0 # 0.00/sec
onblock 2056 # 3.76/sec
cpu-clock 1906902816556 # 1906.903 seconds
task-clock 1906923992083 # 1906.924 seconds
page faults 138839 # 72.808/sec
context switches 34302 # 17.988/sec
cpu migrations 339 # 0.178/sec
major page faults 0 # 0.000/sec
minor page faults 138839 # 72.808/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 773040432866 # 112.899 branches per 1000 inst
branch misses 1708677115 # 0.22% branch miss
conditional 770588283706 # 112.540 conditional branches per 1000 inst
indirect 374106320 # 0.055 indirect branches per 1000 inst
cpu-cycles 1692125537767 # 0.94 GHz
instructions 1370561071471 # 0.81 IPC
slots 3386944277190 #
retiring 409289835857 # 12.1% (18.2%)
-- ucode 463432023 # 0.0%
-- fastpath 408826403834 # 12.1%
frontend 52934629700 # 1.6% ( 2.4%) low
-- latency 26013781506 # 0.8%
-- bandwidth 26920848194 # 0.8%
backend 1774101880707 # 52.4% (79.0%) high
-- cpu 635790079324 # 18.8%
-- memory 1138311801383 # 33.6%
speculation 8381426298 # 0.2% ( 0.4%) low
-- branch mispredict 7898633078 # 0.2%
-- pipeline restart 482793220 # 0.0%
smt-contention 1142234822831 # 33.7% ( 0.0%)
cpu-cycles 1697637019615 # 0.95 GHz
instructions 1372189486239 # 0.81 IPC
instructions 456767926414 # 142.516 l2 access per 1000 inst
l2 hit from l1 26648010394 # 44.94% l2 miss
l2 miss from l1 1513928287 #
l2 hit from l2 pf 10710793952 #
l3 hit from l2 pf 25090146143 #
l3 miss from l2 pf 2647865317 #
instructions 457188883424 # 329.247 float per 1000 inst
float 512 72 # 0.000 AVX-512 per 1000 inst
float 256 344 # 0.000 AVX-256 per 1000 inst
float 128 150528050110 # 329.247 AVX-128 per 1000 inst
float MMX 0 # 0.000 MMX per 1000 inst
float scalar 0 # 0.000 scalar per 1000 inst
instructions 2390641 #
opcache 897366 # 375.366 opcache per 1000 inst
opcache miss 478333 # 53.3% opcache miss rate
l1 dTLB miss 5470 # 2.288 L1 dTLB per 1000 inst
l2 dTLB miss 1094 # 0.458 L2 dTLB per 1000 inst
instructions 2418972 #
icache 1193224 # 493.277 icache per 1000 inst
icache miss 111159 # 9.3% icache miss rate
l1 iTLB miss 7 # 0.003 L1 iTLB per 1000 inst
l2 iTLB miss 0 # 0.000 L2 iTLB per 1000 inst
tlb flush 19 # 0.008 TLB flush per 1000 inst
Intel metrics are much quicker, looks like the AMD version needs multiple runs to reduce tolerance.
elapsed 115.355
on_cpu 0.318 # 5.08 / 16 cores
utime 585.447
stime 0.776
nvcsw 4673 # 48.25%
nivcsw 5012 # 51.75%
inblock 616 # 5.34/sec
onblock 1416 # 12.28/sec
cpu-clock 586340601288 # 586.341 seconds
task-clock 586347075893 # 586.347 seconds
page faults 152353 # 259.834/sec
context switches 10093 # 17.213/sec
cpu migrations 421 # 0.718/sec
major page faults 2 # 0.003/sec
minor page faults 152351 # 259.831/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 157824839681 # 114.070 branches per 1000 inst
branch misses 328274560 # 0.21% branch miss
conditional 157824853057 # 114.070 conditional branches per 1000 inst
indirect 47055770829 # 34.010 indirect branches per 1000 inst
slots 10662556995632 #
retiring 2005207474450 # 18.8% (18.8%)
-- ucode 38968677224 # 0.4%
-- fastpath 1966238797226 # 18.4%
frontend 1142542081923 # 10.7% (10.7%)
-- latency 1027167896973 # 9.6%
-- bandwidth 115374184950 # 1.1%
backend 7477821619753 # 70.1% (70.1%) high
-- cpu 4816740097479 # 45.2%
-- memory 2661081522274 # 25.0%
speculation 51623307331 # 0.5% ( 0.5%) low
-- branch mispredict 48674846500 # 0.5%
-- pipeline restart 2948460831 # 0.0%
smt-contention 0 # 0.0% ( 0.0%)
cpu-cycles 1389467375584 # 0.74 GHz
instructions 1100892371408 # 0.79 IPC
l2 access 104540813869 # 108.996 l2 access per 1000 inst
l2 miss 55139555141 # 52.74% l2 miss
cpu-cycles 1612435730781 # 32.1% memory latency
load stalls 517933116525 # 17.9% l1 bound
l1 miss 229865098888 # 4.7% l2 bound
l2 miss 154331027827 # 5.8% l3 bound
l3 miss 60457495909 # 3.7% dram bound
store_stalls 315269774 # 0.0% store bound
Process overview shows most of the time in clomp-build.
293 processes
48 clomp_build 6106.56 6.88
38 vulkaninfo 1.31 0.95
6 glxinfo:gdrv0 0.14 0.10
4 vulkani:disk$0 0.13 0.10
6 php 0.07 0.05
2 llvmpipe-0 0.07 0.05
2 llvmpipe-1 0.07 0.05
2 llvmpipe-10 0.07 0.05
2 llvmpipe-11 0.07 0.05
2 llvmpipe-12 0.07 0.05
2 llvmpipe-13 0.07 0.05
2 llvmpipe-14 0.07 0.05
2 llvmpipe-15 0.07 0.05
2 llvmpipe-2 0.07 0.05
2 llvmpipe-3 0.07 0.05
2 llvmpipe-4 0.07 0.05
2 llvmpipe-5 0.07 0.05
2 llvmpipe-6 0.07 0.05
2 llvmpipe-7 0.07 0.05
2 llvmpipe-8 0.07 0.05
2 llvmpipe-9 0.07 0.05
2 glxinfo 0.06 0.04
2 glxinfo:cs0 0.06 0.04
2 glxinfo:disk$0 0.06 0.04
2 glxinfo:sh0 0.06 0.04
2 glxinfo:shlo0 0.06 0.04
1 lspci 0.01 0.01
1 ps 0.00 0.01
66 sh 0.00 0.00
13 gcc 0.00 0.00
8 gsettings 0.00 0.00
8 stat 0.00 0.00
8 systemd-detect- 0.00 0.00
5 gmain 0.00 0.00
5 phoronix-test-s 0.00 0.00
3 clomp 0.00 0.00
3 dconf worker 0.00 0.00
2 cc 0.00 0.00
2 lscpu 0.00 0.00
2 uname 0.00 0.00
2 which 0.00 0.00
2 xset 0.00 0.00
1 date 0.00 0.00
1 dirname 0.00 0.00
1 dmesg 0.00 0.00
1 dmidecode 0.00 0.00
1 grep 0.00 0.00
1 ifconfig 0.00 0.00
1 ip 0.00 0.00
1 lsmod 0.00 0.00
1 mktemp 0.00 0.00
1 readlink 0.00 0.00
1 realpath 0.00 0.00
1 sed 0.00 0.00
1 sort 0.00 0.00
1 stty 0.00 0.00
1 systemctl 0.00 0.00
1 template.sh 0.00 0.00
1 wc 0.00 0.00
1 xrandr 0.00 0.00
0 processes running
47 maximum processes
Computation blocks
83118) clomp cpu=3 start=4.90 finish=36.81
83119) clomp_build cpu=13 start=4.90 finish=36.81
83120) clomp_build cpu=15 start=4.90 finish=36.81
83121) clomp_build cpu=0 start=4.90 finish=36.81
83122) clomp_build cpu=14 start=4.90 finish=36.81
83123) clomp_build cpu=4 start=4.90 finish=36.81
83124) clomp_build cpu=9 start=4.90 finish=36.81
83125) clomp_build cpu=10 start=4.90 finish=36.81
83126) clomp_build cpu=3 start=4.90 finish=36.81
83127) clomp_build cpu=5 start=4.90 finish=36.81
83128) clomp_build cpu=1 start=4.90 finish=36.81
83129) clomp_build cpu=7 start=4.90 finish=36.81
83130) clomp_build cpu=8 start=4.90 finish=36.81
83131) clomp_build cpu=12 start=4.90 finish=36.81
83132) clomp_build cpu=6 start=4.90 finish=36.81
83133) clomp_build cpu=2 start=4.90 finish=36.81
83134) clomp_build cpu=11 start=4.90 finish=36.81