This document documented some benchmark results on Apple CPUs.
| GB5 | Geekbench 5 Score | Geekbench 5 Power (watt)¹ | M1 Pro v.s. a14 score | M1 Pro v.s. a14 power | M2 Pro v.s. a15 score | M2 Pro v.s. a15 power |
|---|---|---|---|---|---|---|
| A14 | 1600 | 2.546 | ||||
| M1 Pro | 1780 | 3.244 | 11.3% | 27.4% | ||
| A15 | 1750 | 2.842 | ||||
| M2 | 1890 | 3.767 | ||||
| M2 pro | 1960 | 4.133 | 12.0% | 45.4% |
- ¹ See Notes 2.
| Spec2017 | M1 pro | M2 pro |
|---|---|---|
| 500.perlbench_r | 8.06 | 9.44 |
| 502.gcc_r | 12.60 | 13.91 |
| 505.mcf_r | 8.41 | 9.37 |
| 520.omnetpp_r | 7.80 | 9.73 |
| 523.xalancbmk_r | 9.44 | 10.66 |
| 525.x264_r | 20.00 | 21.74 |
| 531.deepsjeng_r | 5.73 | 6.80 |
| 541.leela_r | 5.77 | 6.43 |
| 548.exchange2_r | 22.30 | 24.27 |
| 557.xz_r | 4.20 | 4.67 |
| 9.10 | 10.32 | |
| Instruction count | ||
| 500.perlbench_r | 2949265681198 | 2952625843135 |
| 502.gcc_r | 1064910378634 | 1067664368104 |
| 505.mcf_r | 920746280386 | 924407291949 |
| 520.omnetpp_r | 1050304739297 | 1052704630840 |
| 523.xalancbmk_r | 1106288805745 | 1106937686333 |
| 525.x264_r | 1511133854481 | 1512998176309 |
| 531.deepsjeng_r | 1753861421785 | 1757013768957 |
| 541.leela_r | 2068151095948 | 2072028906816 |
| 548.exchange2_r | 2132331121002 | 2132678711720 |
| 557.xz_r | 1833513691643 | 1839142287235 |
| 1533687948994.18 | 1536556867013.6 | |
| Branch inst count | ||
| 500.perlbench_r | 624662682256 | 625256513748 |
| 502.gcc_r | 256710738710 | 257190709379 |
| 505.mcf_r | 224623895012 | 225267694175 |
| 520.omnetpp_r | 223164915457 | 223590575917 |
| 523.xalancbmk_r | 322137567978 | 322223831789 |
| 525.x264_r | 106780269260 | 107157785768 |
| 531.deepsjeng_r | 249205164866 | 249751228482 |
| 541.leela_r | 332760649142 | 333465480488 |
| 548.exchange2_r | 285007542905 | 285073548147 |
| 557.xz_r | 316209137574 | 317206891598 |
| 270098618541.075 | 270614170921.339 | |
| Branch miss | ||
| 500.perlbench_r | 2662782515 | 2679148468 |
| 502.gcc_r | 2784039770 | 2781534137 |
| 505.mcf_r | 13112049964 | 13080110376 |
| 520.omnetpp_r | 4662852139 | 4642432136 |
| 523.xalancbmk_r | 633872390 | 626607325 |
| 525.x264_r | 1336374980 | 1327940726 |
| 531.deepsjeng_r | 7060953322 | 6848520890 |
| 541.leela_r | 22132321382 | 22023452519 |
| 548.exchange2_r | 3877629344 | 3853788231 |
| 557.xz_r | 9980440792 | 9946481962 |
| 4330988062.19178 | 4303137107.16249 | |
| Branch inst(%) | ||
| 500.perlbench_r | 21% | 21% |
| 502.gcc_r | 24% | 24% |
| 505.mcf_r | 24% | 24% |
| 520.omnetpp_r | 21% | 21% |
| 523.xalancbmk_r | 29% | 29% |
| 525.x264_r | 7% | 7% |
| 531.deepsjeng_r | 14% | 14% |
| 541.leela_r | 16% | 16% |
| 548.exchange2_r | 13% | 13% |
| 557.xz_r | 17% | 17% |
| 18% | 18% | |
| Branch miss(%) | ||
| 500.perlbench_r | 0.43% | 0.43% |
| 502.gcc_r | 1.08% | 1.08% |
| 505.mcf_r | 5.84% | 5.81% |
| 520.omnetpp_r | 2.09% | 2.08% |
| 523.xalancbmk_r | 0.20% | 0.19% |
| 525.x264_r | 1.25% | 1.24% |
| 531.deepsjeng_r | 2.83% | 2.74% |
| 541.leela_r | 6.65% | 6.60% |
| 548.exchange2_r | 1.36% | 1.35% |
| 557.xz_r | 3.16% | 3.14% |
| 1.60% | 1.59% | |
| Branch MPKI | ||
| 500.perlbench_r | 0.9 | 0.9 |
| 502.gcc_r | 2.6 | 2.6 |
| 505.mcf_r | 14.2 | 14.1 |
| 520.omnetpp_r | 4.4 | 4.4 |
| 523.xalancbmk_r | 0.6 | 0.6 |
| 525.x264_r | 0.9 | 0.9 |
| 531.deepsjeng_r | 4.0 | 3.9 |
| 541.leela_r | 10.7 | 10.6 |
| 548.exchange2_r | 1.8 | 1.8 |
| 557.xz_r | 5.4 | 5.4 |
| 2.8 | 2.8 |
- Compile using gcc12.2 -ofast -march=native Jemalloc 5.3.0
- PMU results are recored via https://gist.github.com/ibireme/173517c208c7dc333ba962c1f0d67d12.
- CPU power(s) are recorded via https://github.com/junjie1475/ios_power_log.git.
- Thread pining support via https://github.com/junjie1475/MacOS_CoreBinder.git.