Embree is set of ray tracing kernels. In the test below, I run three workloads showing slight differences between the first and the other workloads. The workloads are listed as AVX-512 capable, but I don’t see those instructions in my trace. Most likely because I am using default compilation rather than -march=native. An experiment to try in the future. It is floating point intensive code with most time spent in backend memory operations. Branch misprediction is also higher than average.
AMD metrics show a relatively low IPC and many backend misses. The floating point code is predominantly AVX-128 bit.
elapsed 589.246
on_cpu 0.871 # 13.93 / 16 cores
utime 8196.986
stime 11.873
nvcsw 93191 # 41.33%
nivcsw 132272 # 58.67%
inblock 2348544 # 3985.68/sec
onblock 1648 # 2.80/sec
cpu-clock 8210676277740 # 8210.676 seconds
task-clock 8210826881166 # 8210.827 seconds
page faults 2579835 # 314.199/sec
context switches 228219 # 27.795/sec
cpu migrations 657 # 0.080/sec
major page faults 111 # 0.014/sec
minor page faults 2579724 # 314.186/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 1362232904120 # 80.306 branches per 1000 inst
branch misses 184544546458 # 13.55% branch miss
conditional 1140352434370 # 67.226 conditional branches per 1000 inst
indirect 12415374703 # 0.732 indirect branches per 1000 inst
cpu-cycles 34845974472315 # 3.69 GHz
instructions 16962277518169 # 0.49 IPC
slots 69674090415768 #
retiring 7581529448521 # 10.9% (13.5%)
-- ucode 24936273893 # 0.0%
-- fastpath 7556593174628 # 10.8%
frontend 7598757925485 # 10.9% (13.5%)
-- latency 6124859496648 # 8.8%
-- bandwidth 1473898428837 # 2.1%
backend 37121990917073 # 53.3% (65.9%)
-- cpu 9400871625693 # 13.5%
-- memory 27721119291380 # 39.8%
speculation 3987229045557 # 5.7% ( 7.1%)
-- branch mispredict 3967073762620 # 5.7%
-- pipeline restart 20155282937 # 0.0%
smt-contention 13384540352281 # 19.2% ( 0.0%)
cpu-cycles 34862170519282 # 3.68 GHz
instructions 16966704924565 # 0.49 IPC
instructions 5654335743036 # 98.697 l2 access per 1000 inst
l2 hit from l1 440791291703 # 25.36% l2 miss
l2 miss from l1 93098250807 #
l2 hit from l2 pf 68873331326 #
l3 hit from l2 pf 20579962754 #
l3 miss from l2 pf 27819968309 #
instructions 5652446118217 # 296.440 float per 1000 inst
float 512 91 # 0.000 AVX-512 per 1000 inst
float 256 2921024696 # 0.517 AVX-256 per 1000 inst
float 128 1672690446523 # 295.923 AVX-128 per 1000 inst
float MMX 0 # 0.000 MMX per 1000 inst
float scalar 0 # 0.000 scalar per 1000 instIntel metrics including a large percentage of time with branch misses.
elapsed 1463.161
on_cpu 0.904 # 14.46 / 16 cores
utime 21137.635
stime 18.313
nvcsw 368397 # 55.12%
nivcsw 299967 # 44.88%
inblock 0 # 0.00/sec
onblock 1912 # 1.31/sec
cpu-clock 21156464346568 # 21156.464 seconds
task-clock 21156834367225 # 21156.834 seconds
page faults 3295364 # 155.759/sec
context switches 675478 # 31.927/sec
cpu migrations 70820 # 3.347/sec
major page faults 0 # 0.000/sec
minor page faults 3295364 # 155.759/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 3008212425185 # 71.821 branches per 1000 inst
branch misses 441975034062 # 14.69% branch miss
conditional 3008212445665 # 71.821 conditional branches per 1000 inst
indirect 610431103988 # 14.574 indirect branches per 1000 inst
slots 52138940679776 #
retiring 13804850785894 # 26.5% (26.5%)
-- ucode 1354930255474 # 2.6%
-- fastpath 12449920530420 # 23.9%
frontend 8695611248482 # 16.7% (16.7%)
-- latency 5828517798001 # 11.2%
-- bandwidth 2867093450481 # 5.5%
backend 17257445503008 # 33.1% (33.1%)
-- cpu 5308618896422 # 10.2%
-- memory 11948826606586 # 22.9%
speculation 11980383511891 # 23.0% (23.0%)
-- branch mispredict 11929015562360 # 22.9%
-- pipeline restart 51367949531 # 0.1%
smt-contention 0 # 0.0% ( 0.0%)
cpu-cycles 33660387719224 # 2.61 GHz
instructions 26349322565350 # 0.78 IPC
l2 access 824384239858 # 61.635 l2 access per 1000 inst
l2 miss 320741599938 # 38.91% l2 missProcess time is spent in embree_pathtrac program
504 processes
135 embree_pathtrac 123173.29 153.69
64 clinfo 11.20 2.88
38 vulkaninfo 0.75 1.32
6 glxinfo:gdrv0 0.17 0.03
6 php 0.08 0.18
4 vulkani:disk$0 0.08 0.14
2 glxinfo 0.07 0.01
2 glxinfo:cs0 0.07 0.01
2 glxinfo:disk$0 0.07 0.01
2 glxinfo:sh0 0.07 0.01
2 glxinfo:shlo0 0.07 0.01
2 llvmpipe-0 0.04 0.07
2 llvmpipe-1 0.04 0.07
2 llvmpipe-10 0.04 0.07
2 llvmpipe-11 0.04 0.07
2 llvmpipe-12 0.04 0.07
2 llvmpipe-13 0.04 0.07
2 llvmpipe-14 0.04 0.07
2 llvmpipe-15 0.04 0.07
2 llvmpipe-2 0.04 0.07
2 llvmpipe-3 0.04 0.07
2 llvmpipe-4 0.04 0.07
2 llvmpipe-5 0.04 0.07
2 llvmpipe-6 0.04 0.07
2 llvmpipe-7 0.04 0.07
2 llvmpipe-8 0.04 0.07
2 llvmpipe-9 0.04 0.07
6 clang 0.03 0.03
1 lspci 0.00 0.03
92 sh 0.00 0.00
12 gcc 0.00 0.00
10 gsettings 0.00 0.00
9 embree 0.00 0.00
9 stty 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
2 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 ps 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 systemctl 0.00 0.00
1 template.sh 0.00 0.00
1 wc 0.00 0.00
1 xrandr 0.00 0.00
9 processes running
56 maximum processes
Following is the example of the recurring process pattern. Occasionally we miss an “exit” event so that is why 9 were still listed as running above.
6117) embree start=137.17 finish=198.93
6118) embree_pathtrac start=137.17 finish=198.92
6119) embree_pathtrac start=137.37 finish=198.92
6121) embree_pathtrac start=137.37 finish=198.92
6125) embree_pathtrac start=137.37 finish=198.92
6129) embree_pathtrac start=137.37 finish=198.92
6123) embree_pathtrac start=137.37 finish=198.92
6130) embree_pathtrac start=137.37 finish=198.92
6131) embree_pathtrac start=137.37 finish=198.92
6120) embree_pathtrac start=137.37 finish=198.92
6122) embree_pathtrac start=137.37 finish=198.92
6124) embree_pathtrac start=137.37 finish=198.92
6128) embree_pathtrac start=137.37 finish=198.92
6132) embree_pathtrac start=137.37 finish=198.92
6126) embree_pathtrac start=137.37 finish=198.92
6133) embree_pathtrac start=137.37 finish=198.92
6127) embree_pathtrac start=137.37 finish=198.92