quicksilver is a proxy app developed by LLNL. The source is here. There are three workloads. Somehow the AMD processor is much more stable with the workload while the Intel processor needs more iterations to reduce the deviation, particularly on the third workload. The system overview shows full usage of the CPU with a constant set of runnable processes.
The topdown metrics shows a reasonable retirement rate that is limited by backend stalls and with low frontend stalls.
AMD metrics show floating point code with a low branch miss rate and small amount of L2 access.
elapsed 2243.218
on_cpu 0.962 # 15.40 / 16 cores
utime 34535.396
stime 4.859
nvcsw 47991 # 14.52%
nivcsw 282447 # 85.48%
inblock 0 # 0.00/sec
onblock 14152 # 6.31/sec
cpu-clock 34541446980854 # 34541.447 seconds
task-clock 34541661660234 # 34541.662 seconds
page faults 928862 # 26.891/sec
context switches 341458 # 9.885/sec
cpu migrations 748 # 0.022/sec
major page faults 2 # 0.000/sec
minor page faults 928860 # 26.891/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 33807731553674 # 139.022 branches per 1000 inst
branch misses 289158326921 # 0.86% branch miss
conditional 24835473487220 # 102.127 conditional branches per 1000 inst
indirect 93366335290 # 0.384 indirect branches per 1000 inst
cpu-cycles 301810730461036 # 3.82 GHz
instructions 557270879640021 # 1.85 IPC
slots 603566322131106 #
retiring 200786255732776 # 33.3% (53.8%)
-- ucode 119011016871 # 0.0%
-- fastpath 200667244715905 # 33.2%
frontend 25741282012149 # 4.3% ( 6.9%)
-- latency 14193596857908 # 2.4%
-- bandwidth 11547685154241 # 1.9%
backend 131803639022099 # 21.8% (35.3%)
-- cpu 65456513788431 # 10.8%
-- memory 66347125233668 # 11.0%
speculation 15093195866873 # 2.5% ( 4.0%)
-- branch mispredict 14790109025347 # 2.5%
-- pipeline restart 303086841526 # 0.1%
smt-contention 230141589715974 # 38.1% ( 0.0%)
cpu-cycles 136527883836719 # 3.79 GHz
instructions 243228511528062 # 1.78 IPC
instructions 81079749463259 # 22.298 l2 access per 1000 inst
l2 hit from l1 1388992535076 # 6.77% l2 miss
l2 miss from l1 58670581772 #
l2 hit from l2 pf 355183403163 #
l3 hit from l2 pf 54030832873 #
l3 miss from l2 pf 9713216835 #
instructions 81035206592148 # 181.052 float per 1000 inst
float 512 63 # 0.000 AVX-512 per 1000 inst
float 256 600 # 0.000 AVX-256 per 1000 inst
float 128 14671559696429 # 181.052 AVX-128 per 1000 inst
float MMX 0 # 0.000 MMX per 1000 inst
float scalar 5 # 0.000 scalar per 1000 inst
Intel metrics also reflect the longer runtime.
elapsed 5771.167
on_cpu 0.966 # 15.45 / 16 cores
utime 89151.589
stime 8.509
nvcsw 200322 # 25.08%
nivcsw 598411 # 74.92%
inblock 0 # 0.00/sec
onblock 3976 # 0.69/sec
cpu-clock 89157847905491 # 89157.848 seconds
task-clock 89158226755683 # 89158.227 seconds
page faults 1424428 # 15.976/sec
context switches 827367 # 9.280/sec
cpu migrations 60452 # 0.678/sec
major page faults 0 # 0.000/sec
minor page faults 1424428 # 15.976/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 50932134417907 # 143.312 branches per 1000 inst
branch misses 698619389499 # 1.37% branch miss
conditional 50932134437843 # 143.312 conditional branches per 1000 inst
indirect 12618507409368 # 35.506 indirect branches per 1000 inst
slots 440838563740580 #
retiring 173635425896210 # 39.4% (39.4%)
-- ucode 7860615540594 # 1.8%
-- fastpath 165774810355616 # 37.6%
frontend 60186415101796 # 13.7% (13.7%)
-- latency 36475838824133 # 8.3%
-- bandwidth 23710576277663 # 5.4%
backend 168243321796642 # 38.2% (38.2%)
-- cpu 122576607097747 # 27.8%
-- memory 45666714698895 # 10.4%
speculation 39460472909226 # 9.0% ( 9.0%)
-- branch mispredict 39056742849763 # 8.9%
-- pipeline restart 403730059463 # 0.1%
smt-contention 0 # 0.0% ( 0.0%)
cpu-cycles 285841270645200 # 3.06 GHz
instructions 351599636598954 # 1.23 IPC
l2 access 1962502029154 # 11.053 l2 access per 1000 inst
l2 miss 448311834186 # 22.84% l2 miss
Process overview i straightforward
499 processes
144 qs 552276.80 51.29
68 clinfo 15.87 7.65
38 vulkaninfo 0.76 1.52
6 php 0.24 0.12
6 glxinfo:gdrv0 0.12 0.12
4 vulkani:disk$0 0.08 0.16
6 clang 0.07 0.05
2 glxinfo 0.06 0.06
2 glxinfo:cs0 0.06 0.06
2 glxinfo:disk$0 0.06 0.05
2 glxinfo:sh0 0.06 0.04
2 glxinfo:shlo0 0.06 0.04
2 llvmpipe-0 0.04 0.08
2 llvmpipe-1 0.04 0.08
2 llvmpipe-10 0.04 0.08
2 llvmpipe-11 0.04 0.08
2 llvmpipe-12 0.04 0.08
2 llvmpipe-13 0.04 0.08
2 llvmpipe-14 0.04 0.08
2 llvmpipe-15 0.04 0.08
2 llvmpipe-2 0.04 0.08
2 llvmpipe-3 0.04 0.08
2 llvmpipe-4 0.04 0.08
2 llvmpipe-5 0.04 0.08
2 llvmpipe-6 0.04 0.08
2 llvmpipe-7 0.04 0.08
2 llvmpipe-8 0.04 0.08
2 llvmpipe-9 0.04 0.08
1 lspci 0.01 0.02
3 rocminfo 0.00 0.03
1 ps 0.00 0.01
86 sh 0.00 0.00
13 gcc 0.00 0.00
11 gsettings 0.00 0.00
9 quicksilver 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 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 dconf worker 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 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
Following is the compute structure
105523) quicksilver cpu=5 start=6.09 finish=441.82
105524) qs cpu=6 start=6.10 finish=441.82
105525) qs cpu=5 start=6.29 finish=441.82
105526) qs cpu=8 start=6.29 finish=441.82
105527) qs cpu=3 start=6.29 finish=441.82
105528) qs cpu=7 start=6.29 finish=441.82
105529) qs cpu=9 start=6.29 finish=441.82
105530) qs cpu=4 start=6.29 finish=441.82
105531) qs cpu=14 start=6.29 finish=441.82
105532) qs cpu=2 start=6.29 finish=441.82
105533) qs cpu=15 start=6.29 finish=441.82
105534) qs cpu=0 start=6.29 finish=441.82
105535) qs cpu=1 start=6.29 finish=441.82
105536) qs cpu=10 start=6.29 finish=441.82
105537) qs cpu=11 start=6.29 finish=441.82
105538) qs cpu=12 start=6.29 finish=441.82
105539) qs cpu=13 start=6.29 finish=441.82