A package computing discrete fourier transforms. Overall a quick running test.
Topdown profile doesn’t have many data points but shows somewhat high backend stalls.
AMD metrics confirm the backend stalls. The rest of the code has a lot of floating point and low frontend stalls.
elapsed 28.735
on_cpu 0.357 # 5.72 / 16 cores
utime 162.839
stime 1.511
nvcsw 2031 # 48.45%
nivcsw 2161 # 51.55%
inblock 0 # 0.00/sec
onblock 12576 # 437.65/sec
cpu-clock 164440980158 # 164.441 seconds
task-clock 164446588922 # 164.447 seconds
page faults 344700 # 2096.121/sec
context switches 4162 # 25.309/sec
cpu migrations 267 # 1.624/sec
major page faults 2 # 0.012/sec
minor page faults 344698 # 2096.109/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 22525014483 # 29.306 branches per 1000 inst
branch misses 142525357 # 0.63% branch miss
conditional 21315112292 # 27.731 conditional branches per 1000 inst
indirect 46909573 # 0.061 indirect branches per 1000 inst
cpu-cycles 645340177349 # 1.47 GHz
instructions 766939873997 # 1.19 IPC
slots 1295578399398 #
retiring 253149782597 # 19.5% (23.4%)
-- ucode 63790386 # 0.0%
-- fastpath 253085992211 # 19.5%
frontend 18599786075 # 1.4% ( 1.7%) low
-- latency 15794612688 # 1.2%
-- bandwidth 2805173387 # 0.2%
backend 808271511988 # 62.4% (74.8%) high
-- cpu 255543083595 # 19.7%
-- memory 552728428393 # 42.7%
speculation 975977428 # 0.1% ( 0.1%) low
-- branch mispredict 930800943 # 0.1%
-- pipeline restart 45176485 # 0.0%
smt-contention 214580577475 # 16.6% ( 0.0%)
cpu-cycles 643126094843 # 1.49 GHz
instructions 769855395585 # 1.20 IPC
instructions 254848664760 # 26.837 l2 access per 1000 inst
l2 hit from l1 5626317105 # 27.97% l2 miss
l2 miss from l1 1342573404 #
l2 hit from l2 pf 642580883 #
l3 hit from l2 pf 236038076 #
l3 miss from l2 pf 334478930 #
instructions 256599436858 # 606.108 float per 1000 inst
float 512 56 # 0.000 AVX-512 per 1000 inst
float 256 626 # 0.000 AVX-256 per 1000 inst
float 128 155527078471 # 606.108 AVX-128 per 1000 inst
float MMX 0 # 0.000 MMX per 1000 inst
float scalar 0 # 0.000 scalar per 1000 inst
instructions 768061234133 #
opcache 68672979636 # 89.411 opcache per 1000 inst
opcache miss 1051822848 # 1.5% opcache miss rate
l1 dTLB miss 1192723295 # 1.553 L1 dTLB per 1000 inst
l2 dTLB miss 471669429 # 0.614 L2 dTLB per 1000 inst
instructions 768451506345 #
icache 1985334320 # 2.584 icache per 1000 inst
icache miss 244207249 # 12.3% icache miss rate
l1 iTLB miss 9159987 # 0.012 L1 iTLB per 1000 inst
l2 iTLB miss 0 # 0.000 L2 iTLB per 1000 inst
tlb flush 16882 # 0.000 TLB flush per 1000 inst
Intel metrics confirm the dram stalls.
elapsed 24.293
on_cpu 0.389 # 6.22 / 16 cores
utime 149.928
stime 1.105
nvcsw 1443 # 48.36%
nivcsw 1541 # 51.64%
inblock 1584 # 65.20/sec
onblock 1128 # 46.43/sec
cpu-clock 151138846294 # 151.139 seconds
task-clock 151141771739 # 151.142 seconds
page faults 326444 # 2159.853/sec
context switches 2928 # 19.373/sec
cpu migrations 254 # 1.681/sec
major page faults 9 # 0.060/sec
minor page faults 326434 # 2159.787/sec
alignment faults 0 # 0.000/sec
emulation faults 0 # 0.000/sec
branches 22322682452 # 29.078 branches per 1000 inst
branch misses 65479584 # 0.29% branch miss
conditional 22322694676 # 29.078 conditional branches per 1000 inst
indirect 9267166486 # 12.072 indirect branches per 1000 inst
slots 826186570826 #
retiring 379324008162 # 45.9% (45.9%)
-- ucode 4477105158 # 0.5%
-- fastpath 374846903004 # 45.4%
frontend 83216234818 # 10.1% (10.1%)
-- latency 70320222173 # 8.5%
-- bandwidth 12896012645 # 1.6%
backend 358420851091 # 43.4% (43.4%)
-- cpu 169928088458 # 20.6%
-- memory 188492762633 # 22.8%
speculation 5807426002 # 0.7% ( 0.7%) low
-- branch mispredict 5507934937 # 0.7%
-- pipeline restart 299491065 # 0.0%
smt-contention 0 # 0.0% ( 0.0%)
cpu-cycles 272415717150 # 0.70 GHz
instructions 388269199233 # 1.43 IPC
l2 access 8466469743 # 21.811 l2 access per 1000 inst
l2 miss 2813444552 # 33.23% l2 miss
cpu-cycles 272455895543 # 38.5% memory latency
load stalls 92399258775 # 4.7% l1 bound
l1 miss 79638938324 # 5.8% l2 bound
l2 miss 63787703172 # 4.1% l3 bound
l3 miss 52617857012 # 19.3% dram bound
store_stalls 12418643836 # 4.6% store bound
Process summary gives name of the benchmark thread as speed3d
402 processes
48 speed3d 2605.28 9.12
68 clinfo 17.13 5.66
38 vulkaninfo 1.30 1.34
4 vulkani:disk$0 0.13 0.15
6 php 0.08 0.05
2 llvmpipe-0 0.07 0.08
2 llvmpipe-1 0.07 0.08
2 llvmpipe-10 0.07 0.08
2 llvmpipe-11 0.07 0.08
2 llvmpipe-12 0.07 0.08
2 llvmpipe-13 0.07 0.08
2 llvmpipe-14 0.07 0.08
2 llvmpipe-15 0.07 0.08
2 llvmpipe-2 0.07 0.08
2 llvmpipe-3 0.07 0.08
2 llvmpipe-4 0.07 0.08
2 llvmpipe-5 0.07 0.08
2 llvmpipe-6 0.07 0.08
2 llvmpipe-7 0.07 0.08
2 llvmpipe-8 0.07 0.08
2 llvmpipe-9 0.07 0.08
6 glxinfo:gdrv0 0.06 0.13
6 glxinfo:gl0 0.06 0.13
6 clang 0.04 0.08
2 glxinfo 0.04 0.05
2 glxinfo:cs0 0.04 0.05
2 glxinfo:disk$0 0.04 0.05
2 glxinfo:sh0 0.04 0.05
2 glxinfo:shlo0 0.04 0.05
3 rocminfo 0.00 0.03
1 lspci 0.00 0.03
82 sh 0.00 0.00
13 gcc 0.00 0.00
13 gsettings 0.00 0.00
8 stat 0.00 0.00
8 systemd-detect- 0.00 0.00
6 ffte 0.00 0.00
6 llvm-link 0.00 0.00
5 phoronix-test-s 0.00 0.00
2 cc 0.00 0.00
2 gmain 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 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 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
1032593) ffte cpu=9 start=5.82 finish=9.27
1032594) ffte cpu=3 start=5.82 finish=5.82
1032595) speed3d cpu=6 start=5.82 finish=9.26
1032596) speed3d cpu=7 start=5.82 finish=9.26
1032597) speed3d cpu=12 start=5.82 finish=9.26
1032598) speed3d cpu=10 start=5.82 finish=9.26
1032599) speed3d cpu=0 start=5.82 finish=9.26
1032600) speed3d cpu=13 start=5.82 finish=9.26
1032601) speed3d cpu=9 start=5.82 finish=9.26
1032602) speed3d cpu=11 start=5.82 finish=9.26
1032603) speed3d cpu=14 start=5.83 finish=9.26
1032604) speed3d cpu=8 start=5.83 finish=9.26
1032605) speed3d cpu=4 start=5.83 finish=9.26
1032606) speed3d cpu=2 start=5.83 finish=9.26
1032607) speed3d cpu=5 start=5.83 finish=9.26
1032608) speed3d cpu=15 start=5.83 finish=9.26
1032609) speed3d cpu=1 start=5.83 finish=9.26
1032610) speed3d cpu=3 start=5.83 finish=9.26