{"id":1322,"date":"2024-02-03T01:33:01","date_gmt":"2024-02-03T01:33:01","guid":{"rendered":"https:\/\/mvermeulen.org\/perf\/?page_id=1322"},"modified":"2024-02-03T13:21:55","modified_gmt":"2024-02-03T13:21:55","slug":"rnnoise","status":"publish","type":"page","link":"https:\/\/mvermeulen.org\/perf\/workloads\/phoronix\/rnnoise\/","title":{"rendered":"rnnoise"},"content":{"rendered":"\n<p>A recurrent neural network (RNN) being used to remove noise from an audio file. This is a single-threaded test that runs quickly within a minute.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1280\" height=\"960\" src=\"https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/systemtime-14.png\" alt=\"\" class=\"wp-image-1343\" srcset=\"https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/systemtime-14.png 1280w, https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/systemtime-14-1024x768.png 1024w, https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/systemtime-14-768x576.png 768w\" sizes=\"auto, (max-width: 1280px) 100vw, 1280px\" \/><\/figure>\n\n\n\n<p>Topdown profile shows a high retirement rate with some backend stalls.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1280\" height=\"960\" src=\"https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/amdtopdown-14.png\" alt=\"\" class=\"wp-image-1345\" srcset=\"https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/amdtopdown-14.png 1280w, https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/amdtopdown-14-1024x768.png 1024w, https:\/\/mvermeulen.org\/perf\/wp-content\/uploads\/sites\/7\/2024\/02\/amdtopdown-14-768x576.png 768w\" sizes=\"auto, (max-width: 1280px) 100vw, 1280px\" \/><\/figure>\n\n\n\n<p>AMD metrics show this is floating point code, has a lot L2 access rate.  The backend stalls are CPU and not memory. The IPC is also high and the on-code is only one thread.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>elapsed              63.856\non_cpu               0.049          # 0.78 \/ 16 cores\nutime                48.308\nstime                1.258\nnvcsw                2249           # 78.14%\nnivcsw               629            # 21.86%\ninblock              0              # 0.00\/sec\nonblock              830800         # 13010.62\/sec\ncpu-clock            49580676026    # 49.581 seconds\ntask-clock           49583458884    # 49.583 seconds\npage faults          153052         # 3086.755\/sec\ncontext switches     2981           # 60.121\/sec\ncpu migrations       283            # 5.708\/sec\nmajor page faults    2              # 0.040\/sec\nminor page faults    153050         # 3086.715\/sec\nalignment faults     0              # 0.000\/sec\nemulation faults     0              # 0.000\/sec\nbranches             59793040149    # 86.519 branches per 1000 inst\nbranch misses        530140420      # 0.89% branch miss\nconditional          58646138320    # 84.860 conditional branches per 1000 inst\nindirect             73098614       # 0.106 indirect branches per 1000 inst\ncpu-cycles           223501986208   # 0.22 GHz\ninstructions         687431622162   # 3.08 IPC high\nslots                450174913080   #\nretiring             227142696458   # 50.5% (50.5%)\n-- ucode             18076186       #     0.0%\n-- fastpath          227124620272   #    50.5%\nfrontend             38449157594    #  8.5% ( 8.5%)\n-- latency           21249828816    #     4.7%\n-- bandwidth         17199328778    #     3.8%\nbackend              165199650470   # 36.7% (36.7%)\n-- cpu               156259197366   #    34.7%\n-- memory            8940453104     #     2.0%\nspeculation          19315582910    #  4.3% ( 4.3%)\n-- branch mispredict 13384756102    #     3.0%\n-- pipeline restart  5930826808     #     1.3%\nsmt-contention       67545062       #  0.0% ( 0.0%)\ncpu-cycles           223849769349   # 0.22 GHz\ninstructions         688448773654   # 3.08 IPC high\ninstructions         230555151663   # 5.204 l2 access per 1000 inst\nl2 hit from l1       910722479      # 2.37% l2 miss\nl2 miss from l1      16628528       #\nl2 hit from l2 pf    277196298      #\nl3 hit from l2 pf    4811504        #\nl3 miss from l2 pf   7019572        #\ninstructions         230161618213   # 287.307 float per 1000 inst\nfloat 512            60             # 0.000 AVX-512 per 1000 inst\nfloat 256            694            # 0.000 AVX-256 per 1000 inst\nfloat 128            66126936685    # 287.307 AVX-128 per 1000 inst\nfloat MMX            0              # 0.000 MMX per 1000 inst\nfloat scalar         0              # 0.000 scalar per 1000 inst\ninstructions         2658571        #\nopcache              977408         # 367.644 opcache per 1000 inst\nopcache miss         526815         # 53.9% opcache miss rate\nl1 dTLB miss         4675           # 1.758 L1 dTLB per 1000 inst\nl2 dTLB miss         977            # 0.367 L2 dTLB per 1000 inst\ninstructions         2711875        #\nicache               1319384        # 486.521 icache per 1000 inst\nicache miss          110969         #  8.4% icache miss rate\nl1 iTLB miss         13             # 0.005 L1 iTLB per 1000 inst\nl2 iTLB miss         0              # 0.000 L2 iTLB per 1000 inst\ntlb flush            19             # 0.007 TLB flush per 1000 inst\n<\/code><\/pre>\n\n\n\n<p>Intel metrics<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>elapsed              70.816\non_cpu               0.050          # 0.80 \/ 16 cores\nutime                55.772\nstime                0.869\nnvcsw                2614           # 87.22%\nnivcsw               383            # 12.78%\ninblock              912            # 12.88\/sec\nonblock              819560         # 11573.16\/sec\ncpu-clock            56644354100    # 56.644 seconds\ntask-clock           56647446066    # 56.647 seconds\npage faults          142227         # 2510.740\/sec\ncontext switches     3141           # 55.448\/sec\ncpu migrations       317            # 5.596\/sec\nmajor page faults    0              # 0.000\/sec\nminor page faults    142227         # 2510.740\/sec\nalignment faults     0              # 0.000\/sec\nemulation faults     0              # 0.000\/sec\nbranches             59376435112    # 86.107 branches per 1000 inst\nbranch misses        520249709      # 0.88% branch miss\nconditional          59376448392    # 86.107 conditional branches per 1000 inst\nindirect             86449665       # 0.125 indirect branches per 1000 inst\nslots                1277820930584  #\nretiring             697388803314   # 54.6% (54.6%) high\n-- ucode             58189167796    #     4.6%\n-- fastpath          639199635518   #    50.0%\nfrontend             277314084053   # 21.7% (21.7%)\n-- latency           132884455940   #    10.4%\n-- bandwidth         144429628113   #    11.3%\nbackend              164291231386   # 12.9% (12.9%) low\n-- cpu               126996458023   #     9.9%\n-- memory            37294773363    #     2.9%\nspeculation          140493052981   # 11.0% (11.0%) high\n-- branch mispredict 61574068334    #     4.8%\n-- pipeline restart  78918984647    #     6.2%\nsmt-contention       0              #  0.0% ( 0.0%)\ncpu-cycles           213165050359   # 0.19 GHz\ninstructions         689591636128   # 3.24 IPC high\nl2 access            2089755683     # 3.031 l2 access per 1000 inst\nl2 miss              118515140      # 5.67% l2 miss\n<\/code><\/pre>\n\n\n\n<p>Process overview invokes the rnnoise_demo program. It runs fast enough that the test harness includes ~1\/4 of the total user time.<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>396 processes\n\t 30 rnnoise_demo            47.13     0.52\n\t 67 clinfo                  16.34     5.90\n\t 38 vulkaninfo               0.95     1.33\n\t  4 vulkani:disk$0           0.10     0.14\n\t  6 glxinfo:gdrv0            0.08     0.10\n\t  6 glxinfo:gl0              0.08     0.10\n\t  6 clang                    0.06     0.06\n\t  2 llvmpipe-0               0.05     0.07\n\t  2 llvmpipe-1               0.05     0.07\n\t  2 llvmpipe-10              0.05     0.07\n\t  2 llvmpipe-11              0.05     0.07\n\t  2 llvmpipe-12              0.05     0.07\n\t  2 llvmpipe-13              0.05     0.07\n\t  2 llvmpipe-14              0.05     0.07\n\t  2 llvmpipe-15              0.05     0.07\n\t  2 llvmpipe-2               0.05     0.07\n\t  2 llvmpipe-3               0.05     0.07\n\t  2 llvmpipe-4               0.05     0.07\n\t  2 llvmpipe-5               0.05     0.07\n\t  2 llvmpipe-6               0.05     0.07\n\t  2 llvmpipe-7               0.05     0.07\n\t  2 llvmpipe-8               0.05     0.07\n\t  2 llvmpipe-9               0.05     0.07\n\t  2 glxinfo                  0.04     0.04\n\t  2 glxinfo:cs0              0.04     0.04\n\t  2 glxinfo:disk$0           0.04     0.04\n\t  2 glxinfo:sh0              0.04     0.04\n\t  2 glxinfo:shlo0            0.04     0.04\n\t  6 php                      0.03     0.10\n\t  3 rocminfo                 0.03     0.00\n\t  1 lspci                    0.00     0.02\n\t  1 rm                       0.00     0.01\n\t 83 sh                       0.00     0.00\n\t 13 gcc                      0.00     0.00\n\t 12 gsettings                0.00     0.00\n\t 10 sed                      0.00     0.00\n\t  8 stat                     0.00     0.00\n\t  8 systemd-detect-          0.00     0.00\n\t  6 llvm-link                0.00     0.00\n\t  5 phoronix-test-s          0.00     0.00\n\t  3 gmain                    0.00     0.00\n\t  3 ls                       0.00     0.00\n\t  3 rnnoise                  0.00     0.00\n\t  2 cc                       0.00     0.00\n\t  2 lscpu                    0.00     0.00\n\t  2 uname                    0.00     0.00\n\t  2 which                    0.00     0.00\n\t  2 xset                     0.00     0.00\n\t  1 bash                     0.00     0.00\n\t  1 date                     0.00     0.00\n\t  1 dconf worker             0.00     0.00\n\t  1 dirname                  0.00     0.00\n\t  1 dmesg                    0.00     0.00\n\t  1 dmidecode                0.00     0.00\n\t  1 grep                     0.00     0.00\n\t  1 ifconfig                 0.00     0.00\n\t  1 ip                       0.00     0.00\n\t  1 lsmod                    0.00     0.00\n\t  1 mktemp                   0.00     0.00\n\t  1 ps                       0.00     0.00\n\t  1 qdbus                    0.00     0.00\n\t  1 readlink                 0.00     0.00\n\t  1 realpath                 0.00     0.00\n\t  1 sort                     0.00     0.00\n\t  1 stty                     0.00     0.00\n\t  1 systemctl                0.00     0.00\n\t  1 template.sh              0.00     0.00\n\t  1 wc                       0.00     0.00\n\t  1 xrandr                   0.00     0.00\n0 processes running\n47 maximum processes\n<\/code><\/pre>\n\n\n\n<p>Computation structure<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>      204018) rnnoise          cpu=0 start=5.59  finish=21.43\n        204019) rnnoise_demo     cpu=3 start=5.59  finish=21.43\n          204020) rnnoise_demo     cpu=15 start=5.59  finish=5.59 \n            204021) rnnoise_demo     cpu=12 start=5.59  finish=5.59 \n          204022) rnnoise_demo     cpu=5 start=5.59  finish=5.59 \n          204023) rnnoise_demo     cpu=2 start=5.59  finish=5.60 \n            204024) rnnoise_demo     cpu=1 start=5.59  finish=5.59 \n            204025) sed              cpu=14 start=5.60  finish=5.60 \n          204026) rnnoise_demo     cpu=12 start=5.60  finish=5.60 \n            204027) ls               cpu=5 start=5.60  finish=5.60 \n            204028) sed              cpu=15 start=5.60  finish=5.60 \n          204029) rnnoise_demo     cpu=9 start=5.60  finish=5.60 \n          204030) rnnoise_demo     cpu=2 start=5.60  finish=5.61 \n            204031) rnnoise_demo     cpu=14 start=5.61  finish=5.61 \n            204032) sed              cpu=0 start=5.61  finish=5.61 \n<\/code><\/pre>\n","protected":false},"excerpt":{"rendered":"<p>A recurrent neural network (RNN) being used to remove noise from an audio file. This is a single-threaded test that runs quickly within a minute. Topdown profile shows a high retirement rate with some backend stalls. AMD metrics show this <span class=\"excerpt-dots\">&hellip;<\/span> <a class=\"more-link\" href=\"https:\/\/mvermeulen.org\/perf\/workloads\/phoronix\/rnnoise\/\"><span class=\"more-msg\">Continue reading &rarr;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"parent":58,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-1322","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/1322","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/comments?post=1322"}],"version-history":[{"count":3,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/1322\/revisions"}],"predecessor-version":[{"id":1346,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/1322\/revisions\/1346"}],"up":[{"embeddable":true,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/58"}],"wp:attachment":[{"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/media?parent=1322"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}