{"id":2297,"date":"2024-06-01T13:51:20","date_gmt":"2024-06-01T13:51:20","guid":{"rendered":"https:\/\/mvermeulen.org\/perf\/?page_id=2297"},"modified":"2024-06-05T12:17:09","modified_gmt":"2024-06-05T12:17:09","slug":"cpu2017","status":"publish","type":"page","link":"https:\/\/mvermeulen.org\/perf\/workloads\/cpu2017\/","title":{"rendered":"cpu2017"},"content":{"rendered":"\n

CPU 2017<\/a> is an industry benchmark used to evaluate compute-intensive workloads and in particular a focus on the processor, memory subsystem and compiler. Somewhat expensive ($1000) but otherwise interesting in evaluation\/comparison as workloads.<\/p>\n\n\n\n

There is a more formal set of rules that govern running of the tests and disclosure of the results. My focus is more on understanding the workloads than creating formal submittals, so anything here should be considered a “estimate” that doesn’t follow strict reporting, isn’t trying to maximize absolute results and is not a disclosure. At the same time, characterizing unoptimized code is also not interesting, so I did use a config file with the AMD AOCC compiler <\/a>and many common “base” options and run on 7840 processor. For example, performance measurements include elapsed time running the test harness and not just the benchmark executables.<\/p>\n\n\n\n

More detailed descriptions of the benchmarks can be found here<\/a>. Most of my interest is on the rate benchmarks, e.g. intrate and fprate.<\/p>\n\n\n\n\n\n\n\t\n\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t\n\t
benchmark<\/th>status<\/th>elapsed<\/th>on_cpu<\/th>inblock<\/th>onblock<\/th>page-fault<\/th>context-switch<\/th>IPC<\/th>GHz<\/th>retire-rate<\/th>frontend-stall<\/th>backend-stall<\/th>spec-stall<\/th>retire-ucode<\/th>retire-fastpath<\/th>float-density<\/th>frontend-latency<\/th>frontend-bandwidth<\/th>opcache-miss<\/th>icache-miss<\/th>backend-cpu<\/th>backend-memory<\/th>amd-l2-miss<\/th>amd-l2-density<\/th>spec-branch<\/th>spec-pipeline<\/th>branch-miss<\/th>branch-density<\/th>branch-cond<\/th>branch-ind<\/th>smt-contention<\/th>\n<\/tr>\n<\/thead>\n
500.perlbench_r<\/td>CPU<\/td>1272.531<\/td>15.75<\/td>0.02<\/td>542.18<\/td>389.437<\/td>10.721<\/td>1.58<\/td>4.10<\/td>33.3<\/td>19.7<\/td>44.7<\/td>2.3<\/td>0.1<\/td>25.2<\/td>16.803<\/td>8.8<\/td>6.2<\/td>5.6<\/td>42.5<\/td>3.1<\/td>30.8<\/td>11.62<\/td>15.736<\/td>1.7<\/td>0.1<\/td>0.63<\/td>184.569<\/td>132.594<\/td>12.996<\/td>23.9<\/td>\n<\/tr>\n
502.gcc_r<\/td>CPU<\/td>1280.597<\/td>15.70<\/td>0.00<\/td>4845.61<\/td>26600.971<\/td>10.948<\/td>0.60<\/td>4.39<\/td>12.7<\/td>25.6<\/td>59.8<\/td>1.9<\/td>0.0<\/td>9.6<\/td>7.943<\/td>12.6<\/td>6.9<\/td>15.2<\/td>29.7<\/td>3.1<\/td>42.5<\/td>29.37<\/td>60.271<\/td>1.4<\/td>0.0<\/td>1.99<\/td>222.751<\/td>170.047<\/td>5.594<\/td>23.8<\/td>\n<\/tr>\n
503.bwaves_r<\/td>CPU<\/td>4632.332<\/td>15.85<\/td>0.00<\/td>4.72<\/td>565.445<\/td>11.053<\/td>0.13<\/td>4.47<\/td>2.3<\/td>1.4<\/td>96.3<\/td>0.1<\/td>0.0<\/td>2.2<\/td>260.237<\/td>1.2<\/td>0.2<\/td>9.7<\/td>6.5<\/td>6.8<\/td>86.5<\/td>49.20<\/td>132.988<\/td>0.0<\/td>0.0<\/td>1.38<\/td>19.172<\/td>14.975<\/td>1.398<\/td>3.0<\/td>\n<\/tr>\n
505.mcf_r<\/td>CPU<\/td>706.488<\/td>15.73<\/td>0.00<\/td>15.39<\/td>1495.202<\/td>10.497<\/td>0.98<\/td>4.06<\/td>19.9<\/td>26.9<\/td>43.9<\/td>9.3<\/td>0.0<\/td>16.3<\/td>0.687<\/td>13.1<\/td>8.9<\/td>0.4<\/td>16.7<\/td>4.3<\/td>31.6<\/td>16.33<\/td>53.401<\/td>7.6<\/td>0.0<\/td>4.98<\/td>169.000<\/td>147.312<\/td>0.017<\/td>18.2<\/td>\n<\/tr>\n
507.cactuBSSN_r<\/td>CPU<\/td>609.922<\/td>15.69<\/td>0.00<\/td>495.30<\/td>1135.260<\/td>13.491<\/td>0.23<\/td>4.21<\/td>4.3<\/td>4.8<\/td>90.9<\/td>0.1<\/td>0.0<\/td>4.1<\/td>55.472<\/td>3.6<\/td>1.1<\/td>47.7<\/td>41.8<\/td>9.1<\/td>78.5<\/td>10.76<\/td>476.467<\/td>0.0<\/td>0.0<\/td>0.64<\/td>48.787<\/td>33.403<\/td>3.734<\/td>3.6<\/td>\n<\/tr>\n
508.namd_r<\/td>CPU<\/td>705.983<\/td>15.81<\/td>0.00<\/td>45.01<\/td>239.918<\/td>11.385<\/td>1.83<\/td>3.74<\/td>51.3<\/td>4.8<\/td>40.9<\/td>3.0<\/td>0.0<\/td>31.3<\/td>395.622<\/td>2.2<\/td>0.8<\/td>0.2<\/td>18.8<\/td>17.2<\/td>7.7<\/td>1.35<\/td>63.010<\/td>1.8<\/td>0.0<\/td>4.40<\/td>26.587<\/td>24.258<\/td>0.021<\/td>39.1<\/td>\n<\/tr>\n
510.parest_r<\/td>CPU<\/td>4550.517<\/td>15.80<\/td>0.00<\/td>16.88<\/td>213.125<\/td>10.974<\/td>0.29<\/td>4.41<\/td>5.8<\/td>4.2<\/td>89.7<\/td>0.3<\/td>0.0<\/td>4.8<\/td>338.504<\/td>1.9<\/td>1.6<\/td>1.4<\/td>23.3<\/td>6.8<\/td>67.4<\/td>34.77<\/td>93.116<\/td>0.2<\/td>0.0<\/td>1.59<\/td>106.362<\/td>90.964<\/td>3.249<\/td>17.3<\/td>\n<\/tr>\n
511.povray_r<\/td>CPU<\/td>1207.797<\/td>15.60<\/td>0.00<\/td>1449.68<\/td>63.683<\/td>10.402<\/td>1.81<\/td>3.78<\/td>52.4<\/td>6.6<\/td>38.9<\/td>2.1<\/td>0.3<\/td>31.4<\/td>244.832<\/td>2.5<\/td>1.5<\/td>3.7<\/td>20.9<\/td>11.2<\/td>12.4<\/td>0.06<\/td>63.187<\/td>1.1<\/td>0.2<\/td>0.25<\/td>157.689<\/td>109.665<\/td>10.900<\/td>39.5<\/td>\n<\/tr>\n
519.lbm_r<\/td>CPU<\/td>1416.599<\/td>15.53<\/td>0.00<\/td>7.13<\/td>256.969<\/td>10.588<\/td>0.26<\/td>4.42<\/td>4.7<\/td>2.1<\/td>93.2<\/td>0.0<\/td>0.0<\/td>4.5<\/td>51.354<\/td>1.1<\/td>1.0<\/td>15.3<\/td>6.0<\/td>2.2<\/td>88.9<\/td>24.72<\/td>172.647<\/td>0.0<\/td>0.0<\/td>0.08<\/td>138.203<\/td>137.728<\/td>0.026<\/td>2.3<\/td>\n<\/tr>\n
520.omnetpp_r<\/td>CPU<\/td>2024.583<\/td>15.83<\/td>0.00<\/td>870.95<\/td>107.720<\/td>9.994<\/td>0.28<\/td>4.58<\/td>5.6<\/td>9.6<\/td>81.9<\/td>3.0<\/td>0.0<\/td>4.9<\/td>16.375<\/td>4.4<\/td>3.9<\/td>2.3<\/td>28.4<\/td>4.8<\/td>66.7<\/td>44.62<\/td>72.990<\/td>2.5<\/td>0.1<\/td>3.02<\/td>196.963<\/td>143.767<\/td>11.785<\/td>12.8<\/td>\n<\/tr>\n
521.wrf_r<\/td>CPU<\/td>1953.561<\/td>15.79<\/td>0.00<\/td>1454.42<\/td>198.479<\/td>12.169<\/td>0.43<\/td>4.35<\/td>7.8<\/td>7.9<\/td>83.8<\/td>0.5<\/td>0.0<\/td>7.4<\/td>278.072<\/td>6.2<\/td>1.2<\/td>3.1<\/td>27.3<\/td>15.4<\/td>63.5<\/td>26.15<\/td>77.326<\/td>0.4<\/td>0.0<\/td>0.90<\/td>113.379<\/td>77.677<\/td>13.945<\/td>5.9<\/td>\n<\/tr>\n
523.xalancbmk_r<\/td>CPU<\/td>716.422<\/td>15.72<\/td>0.00<\/td>8293.00<\/td>930.346<\/td>10.489<\/td>0.77<\/td>4.32<\/td>19.0<\/td>8.9<\/td>71.1<\/td>0.9<\/td>0.1<\/td>12.2<\/td>34.341<\/td>3.6<\/td>2.2<\/td>4.0<\/td>20.7<\/td>3.0<\/td>42.9<\/td>15.96<\/td>75.229<\/td>0.5<\/td>0.1<\/td>0.34<\/td>267.659<\/td>234.583<\/td>7.449<\/td>35.4<\/td>\n<\/tr>\n
525.x264_r<\/td>CPU<\/td>427.195<\/td>15.00<\/td>0.00<\/td>16419.20<\/td>1583.996<\/td>12.872<\/td>1.59<\/td>3.69<\/td>38.0<\/td>21.0<\/td>38.6<\/td>2.5<\/td>0.1<\/td>27.2<\/td>189.560<\/td>11.2<\/td>3.9<\/td>28.6<\/td>42.4<\/td>11.4<\/td>16.3<\/td>6.16<\/td>37.180<\/td>1.8<\/td>0.0<\/td>2.76<\/td>66.296<\/td>48.490<\/td>3.845<\/td>28.0<\/td>\n<\/tr>\n
526.blender_r<\/td>CPU<\/td>1003.543<\/td>15.62<\/td>0.00<\/td>655.15<\/td>1981.576<\/td>11.564<\/td>1.08<\/td>3.88<\/td>24.9<\/td>6.6<\/td>62.3<\/td>6.2<\/td>0.0<\/td>17.9<\/td>397.608<\/td>3.7<\/td>1.1<\/td>1.0<\/td>28.3<\/td>13.6<\/td>31.2<\/td>15.56<\/td>31.225<\/td>4.5<\/td>0.0<\/td>2.09<\/td>134.627<\/td>120.414<\/td>1.256<\/td>28.1<\/td>\n<\/tr>\n
527.cam4_r<\/td>CPU<\/td>1312.508<\/td>15.81<\/td>7.50<\/td>935.89<\/td>3656.413<\/td>10.919<\/td>0.84<\/td>4.07<\/td>16.4<\/td>16.0<\/td>66.9<\/td>0.7<\/td>0.0<\/td>14.5<\/td>189.588<\/td>9.3<\/td>4.9<\/td>8.0<\/td>25.9<\/td>19.5<\/td>40.0<\/td>20.78<\/td>62.695<\/td>0.6<\/td>0.0<\/td>1.15<\/td>124.383<\/td>89.321<\/td>8.772<\/td>11.1<\/td>\n<\/tr>\n
531.deepsjeng_r<\/td>CPU<\/td>811.223<\/td>15.78<\/td>0.00<\/td>37.06<\/td>718.782<\/td>10.181<\/td>1.42<\/td>4.08<\/td>30.2<\/td>29.4<\/td>34.9<\/td>5.5<\/td>0.0<\/td>23.6<\/td>21.274<\/td>15.0<\/td>7.9<\/td>17.7<\/td>16.4<\/td>3.7<\/td>23.6<\/td>4.85<\/td>23.537<\/td>4.2<\/td>0.1<\/td>3.99<\/td>123.840<\/td>97.520<\/td>0.907<\/td>21.9<\/td>\n<\/tr>\n
538.imagick_r<\/td>CPU<\/td>371.223<\/td>14.33<\/td>0.00<\/td>1329.38<\/td>1026.201<\/td>10.757<\/td>2.19<\/td>3.63<\/td>56.0<\/td>11.1<\/td>25.2<\/td>7.7<\/td>0.0<\/td>33.8<\/td>149.855<\/td>3.8<\/td>2.9<\/td>0.1<\/td>14.4<\/td>8.3<\/td>6.9<\/td>6.13<\/td>11.978<\/td>4.6<\/td>0.0<\/td>0.89<\/td>182.276<\/td>175.161<\/td>0.187<\/td>39.6<\/td>\n<\/tr>\n
541.leela_r<\/td>CPU<\/td>1062.974<\/td>15.76<\/td>0.00<\/td>144.03<\/td>156.446<\/td>10.467<\/td>1.05<\/td>4.07<\/td>23.6<\/td>50.0<\/td>12.7<\/td>13.7<\/td>0.0<\/td>18.2<\/td>81.302<\/td>27.5<\/td>11.2<\/td>2.6<\/td>8.3<\/td>3.0<\/td>6.8<\/td>4.14<\/td>18.021<\/td>10.5<\/td>0.1<\/td>12.17<\/td>141.333<\/td>118.833<\/td>0.187<\/td>22.7<\/td>\n<\/tr>\n
544.nab_r<\/td>CPU<\/td>581.066<\/td>15.53<\/td>0.00<\/td>23.20<\/td>301.419<\/td>11.174<\/td>1.27<\/td>3.77<\/td>36.9<\/td>8.7<\/td>52.1<\/td>2.3<\/td>0.1<\/td>25.5<\/td>318.530<\/td>5.0<\/td>1.1<\/td>1.4<\/td>15.1<\/td>22.8<\/td>13.3<\/td>4.94<\/td>52.578<\/td>1.5<\/td>0.0<\/td>1.30<\/td>83.289<\/td>72.175<\/td>1.774<\/td>30.8<\/td>\n<\/tr>\n
548.exchange2_r<\/td>CPU<\/td>557.819<\/td>15.77<\/td>0.00<\/td>23.76<\/td>71.047<\/td>10.670<\/td>1.89<\/td>3.99<\/td>46.4<\/td>36.9<\/td>14.2<\/td>2.6<\/td>0.0<\/td>32.3<\/td>126.182<\/td>12.0<\/td>13.6<\/td>1.6<\/td>17.7<\/td>4.6<\/td>5.3<\/td>0.76<\/td>0.827<\/td>1.8<\/td>0.0<\/td>1.30<\/td>165.361<\/td>157.689<\/td>1.022<\/td>30.3<\/td>\n<\/tr>\n
549.fotonik3d_r<\/td>CPU<\/td>4829.371<\/td>15.91<\/td>0.00<\/td>32.94<\/td>144.718<\/td>11.510<\/td>0.11<\/td>4.54<\/td>2.0<\/td>1.9<\/td>96.1<\/td>0.1<\/td>0.0<\/td>1.9<\/td>286.386<\/td>1.3<\/td>0.5<\/td>6.8<\/td>12.8<\/td>2.7<\/td>91.8<\/td>44.60<\/td>137.645<\/td>0.0<\/td>0.0<\/td>0.45<\/td>36.518<\/td>33.907<\/td>0.471<\/td>1.7<\/td>\n<\/tr>\n
554.roms_r<\/td>CPU<\/td>2913.604<\/td>15.82<\/td>0.00<\/td>8.12<\/td>600.297<\/td>11.852<\/td>0.16<\/td>4.45<\/td>2.8<\/td>2.3<\/td>94.9<\/td>0.1<\/td>0.0<\/td>2.8<\/td>129.670<\/td>1.8<\/td>0.4<\/td>4.4<\/td>21.7<\/td>7.4<\/td>85.6<\/td>36.36<\/td>196.455<\/td>0.0<\/td>0.0<\/td>0.43<\/td>76.566<\/td>57.327<\/td>6.664<\/td>2.0<\/td>\n<\/tr>\n
557.xz_r<\/td>CPU<\/td>1413.262<\/td>15.76<\/td>0.00<\/td>11.48<\/td>1363.376<\/td>9.484<\/td>0.77<\/td>4.41<\/td>19.4<\/td>9.5<\/td>64.8<\/td>6.3<\/td>0.0<\/td>12.6<\/td>21.356<\/td>4.5<\/td>1.7<\/td>1.0<\/td>17.8<\/td>2.2<\/td>39.8<\/td>30.11<\/td>23.253<\/td>4.1<\/td>0.0<\/td>4.53<\/td>115.607<\/td>104.940<\/td>1.340<\/td>35.2<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n","protected":false},"excerpt":{"rendered":"

CPU 2017 is an industry benchmark used to evaluate compute-intensive workloads and in particular a focus on the processor, memory subsystem and compiler. Somewhat expensive ($1000) but otherwise interesting in evaluation\/comparison as workloads. There is a more formal set of …<\/span> Continue reading →<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"parent":48,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-2297","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/2297","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=2297"}],"version-history":[{"count":2,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/2297\/revisions"}],"predecessor-version":[{"id":2434,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/2297\/revisions\/2434"}],"up":[{"embeddable":true,"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/pages\/48"}],"wp:attachment":[{"href":"https:\/\/mvermeulen.org\/perf\/wp-json\/wp\/v2\/media?parent=2297"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}