A neural network framework with 24 different workloads. Mostly runs on half of the logical cores.
Topdown metrics show a backend bound workload but occasional front-end stalls.
AMD topdown metrics show a small amount of floating point and not many branches. The backend stalls are at least as much CPU as memory.
elapsed 3569.478
on_cpu 0.326 # 5.22 / 16 cores
utime 18324.364
stime 307.259
nvcsw 11222907 # 99.54%
nivcsw 52164 # 0.46%
inblock 1228440 # 344.15/sec
onblock 42214160 # 11826.42/sec
cpu-clock 18625705808601 # 18625.706 seconds
task-clock 18629500805795 # 18629.501 seconds
page faults 69392523 # 3724.873/sec
context switches 11292506 # 606.163/sec
cpu migrations 5906 # 0.317/sec
major page faults 243 # 0.013/sec
minor page faults 69392280 # 3724.860/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 3840677894116 # 42.361 branches per 1000 inst
branch misses 101209872829 # 2.64% branch miss
conditional 2983734079416 # 32.909 conditional branches per 1000 inst
indirect 118343443749 # 1.305 indirect branches per 1000 inst
cpu-cycles 78591628660514 # 1.38 GHz
instructions 90362073448677 # 1.15 IPC
slots 157193781261678 #
retiring 33408779346212 # 21.3% (21.3%)
-- ucode 107489339676 # 0.1%
-- fastpath 33301290006536 # 21.2%
frontend 14717913722454 # 9.4% ( 9.4%)
-- latency 11303657501748 # 7.2%
-- bandwidth 3414256220706 # 2.2%
backend 107252745611847 # 68.2% (68.3%)
-- cpu 62638564249864 # 39.8%
-- memory 44614181361983 # 28.4%
speculation 1668870353449 # 1.1% ( 1.1%)
-- branch mispredict 1552840004870 # 1.0%
-- pipeline restart 116030348579 # 0.1%
smt-contention 145376168253 # 0.1% ( 0.0%)
cpu-cycles 78661479112363 # 1.38 GHz
instructions 90267718612366 # 1.15 IPC
instructions 30088349387486 # 156.378 l2 access per 1000 inst
l2 hit from l1 3522876544780 # 11.60% l2 miss
l2 miss from l1 250597431430 #
l2 hit from l2 pf 887065395908 #
l3 hit from l2 pf 228456679605 #
l3 miss from l2 pf 66743478461 #
instructions 30088835388315 # 50.472 float per 1000 inst
float 512 133 # 0.000 AVX-512 per 1000 inst
float 256 381138356114 # 12.667 AVX-256 per 1000 inst
float 128 1137465319616 # 37.804 AVX-128 per 1000 inst
float MMX 0 # 0.000 MMX per 1000 inst
float scalar 29596531 # 0.001 scalar per 1000 inst
Intel metrics show ~3 of 16 cores.
elapsed 3686.610
on_cpu 0.189 # 3.03 / 16 cores
utime 10969.732
stime 203.687
nvcsw 6807043 # 98.55%
nivcsw 100101 # 1.45%
inblock 1998880 # 542.20/sec
onblock 48694912 # 13208.59/sec
cpu-clock 11168917301443 # 11168.917 seconds
task-clock 11169935020733 # 11169.935 seconds
page faults 69133490 # 6189.247/sec
context switches 6925161 # 619.982/sec
cpu migrations 4705 # 0.421/sec
major page faults 3716 # 0.333/sec
minor page faults 69129774 # 6188.915/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 4876027591344 # 42.096 branches per 1000 inst
branch misses 45054577411 # 0.92% branch miss
conditional 4876027683984 # 42.096 conditional branches per 1000 inst
indirect 132760442627 # 1.146 indirect branches per 1000 inst
slots 249817192973432 #
retiring 113578962480686 # 45.5% (45.5%)
-- ucode 4381356017946 # 1.8%
-- fastpath 109197606462740 # 43.7%
frontend 15248899292191 # 6.1% ( 6.1%)
-- latency 9155034273169 # 3.7%
-- bandwidth 6093865019022 # 2.4%
backend 115519281492570 # 46.2% (46.2%)
-- cpu 81476276360751 # 32.6%
-- memory 34043005131819 # 13.6%
speculation 6054177131575 # 2.4% ( 2.4%)
-- branch mispredict 5526964401535 # 2.2%
-- pipeline restart 527212730040 # 0.2%
smt-contention 0 # 0.0% ( 0.0%)
cpu-cycles 41759566746190 # 0.72 GHz
instructions 115237922937775 # 2.76 IPC
l2 access 6658511399645 # 57.800 l2 access per 1000 inst
l2 miss 1068977335236 # 16.05% l2 miss
Process overview shows an internal benchmark process
1637 processes
1152 deepsparse.benc 517.65 1067.38
68 clinfo 19.83 6.33
38 vulkaninfo 0.97 1.53
6 glxinfo:gdrv0 0.19 0.09
6 php 0.16 0.40
4 vulkani:disk$0 0.11 0.17
2 glxinfo 0.09 0.03
2 glxinfo:cs0 0.09 0.03
2 glxinfo:disk$0 0.09 0.03
2 glxinfo:sh0 0.09 0.03
2 glxinfo:shlo0 0.09 0.03
2 llvmpipe-0 0.06 0.09
2 llvmpipe-1 0.06 0.09
2 llvmpipe-10 0.06 0.09
2 llvmpipe-11 0.06 0.09
2 llvmpipe-12 0.06 0.09
2 llvmpipe-13 0.06 0.09
2 llvmpipe-14 0.06 0.09
2 llvmpipe-15 0.06 0.09
2 llvmpipe-3 0.06 0.09
2 llvmpipe-4 0.06 0.09
2 llvmpipe-5 0.06 0.09
2 llvmpipe-6 0.06 0.09
2 llvmpipe-7 0.06 0.09
2 llvmpipe-8 0.06 0.09
2 llvmpipe-9 0.06 0.09
2 llvmpipe-2 0.06 0.08
6 clang 0.05 0.07
3 rocminfo 0.03 0.00
1 lspci 0.01 0.02
1 ps 0.00 0.01
81 sh 0.00 0.00
72 arch.bin 0.00 0.00
72 deepsparse 0.00 0.00
12 gcc 0.00 0.00
9 gsettings 0.00 0.00
8 stat 0.00 0.00
8 systemd-detect- 0.00 0.00
6 llvm-link 0.00 0.00
5 phoronix-test-s 0.00 0.00
4 gmain 0.00 0.00
3 dconf worker 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 cc 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 python 0.00 0.00
1 python3 0.00 0.00
1 qdbus 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
Regular set of processes started
2648224) deepsparse cpu=14 start=5.84 finish=7.55
2648225) deepsparse.benc cpu=15 start=5.85 finish=7.54
2648226) deepsparse.benc cpu=11 start=5.88 finish=7.54
2648227) deepsparse.benc cpu=5 start=5.88 finish=7.54
2648228) deepsparse.benc cpu=9 start=5.88 finish=7.54
2648229) deepsparse.benc cpu=10 start=5.88 finish=7.54
2648230) deepsparse.benc cpu=12 start=5.88 finish=7.54
2648231) deepsparse.benc cpu=14 start=5.88 finish=7.54
2648232) deepsparse.benc cpu=8 start=5.88 finish=7.54
2648233) deepsparse.benc cpu=0 start=5.88 finish=7.54
2648234) deepsparse.benc cpu=1 start=5.88 finish=7.54
2648235) deepsparse.benc cpu=3 start=5.88 finish=7.54
2648236) deepsparse.benc cpu=13 start=5.88 finish=7.54
2648237) deepsparse.benc cpu=2 start=5.88 finish=7.54
2648238) deepsparse.benc cpu=6 start=5.88 finish=7.54
2648239) deepsparse.benc cpu=4 start=5.88 finish=7.54
2648240) deepsparse.benc cpu=7 start=5.88 finish=7.54
2648241) arch.bin cpu=15 start=7.52 finish=7.53