3 `powermetrics` is a Mac-only command-line utility that provides many
4 high-quality power-related measurements. It is most useful for getting
5 CPU, GPU and wakeup measurements in a precise and easily scriptable
6 fashion (unlike [Activity Monitor and
7 top](activity_monitor_and_top.md))
8 especially in combination with
9 [rapl](tools_power_rapl.md) via the
10 `mach power` command. This document describes the version of
11 `powermetrics` that comes with Mac OS 10.10. The one that comes with
12 10.9 is less powerful.
14 **Note**: The [power profiling
15 overview](power_profiling_overview.md) is
16 worth reading at this point if you haven\'t already. It may make parts
17 of this document easier to understand.
21 `powermetrics` provides a vast number of measurements. The following
22 command encompasses the most useful ones:
24 sudo powermetrics --samplers tasks --show-process-coalition --show-process-gpu -n 1 -i 5000
26 - `--samplers tasks` tells it to just do per-process measurements.
27 - `--show-process-coalition`` `tells it to group *coalitions* of
28 related processes, e.g. the Firefox parent process and child
30 - `--show-process-gpu` tells it to show per-process GPU measurements.
31 - `-n 1` tells it to take one sample and then stop.
32 - `-i 5000` tells it to use a sample length of 5 seconds (5000 ms).
33 Change this number to get shorter or longer samples.
35 The following is example output from such an invocation:
37 *** Sampled system activity (Fri Sep 4 17:15:14 2015 +1000) (5009.63ms elapsed) ***
41 Name ID CPU ms/s User% Deadlines (<2 ms, 2-5 ms) Wakeups (Intr, Pkg idle) GPU ms/s
42 com.apple.Terminal 293 447.66 274.83 120.35 221.74
43 firefox 84627 77.59 55.55 15.37 2.59 91.42 42.12 204.47
44 plugin-container 84628 377.22 37.18 43.91 18.56 178.65 75.85 17.29
45 Terminal 694 9.86 79.94 0.00 0.00 4.39 2.20 0.00
46 powermetrics 84694 1.21 31.53 0.00 0.00 0.20 0.20 0.00
47 com.google.Chrome 489 233.83 48.10 25.95 0.00
48 Google Chrome Helper 84688 181.57 92.81 0.00 0.00 23.95 12.77 0.00
49 Google Chrome 84681 57.26 76.07 4.39 0.00 23.75 12.97 0.00
50 Google Chrome Helper 84685 0.13 48.08 0.00 0.00 0.40 0.20 0.00
51 kernel_coalition 1 128.64 780.19 330.52 0.00
52 kernel_task 0 109.97 0.00 0.20 0.00 779.47 330.35 0.00
53 launchd 1 18.88 2.44 0.00 0.00 0.40 0.20 0.00
54 com.apple.Safari 488 90.60 108.58 56.48 26.65
55 com.apple.WebKit.WebContent 84679 64.21 84.69 0.00 0.00 104.19 54.89 26.66
56 com.apple.WebKit.Networking 84678 26.89 58.89 0.40 0.00 1.60 0.00 0.00
57 Safari 84676 1.56 55.74 0.00 0.00 2.59 1.40 0.00
58 com.apple.Safari.SearchHelper 84690 0.15 49.49 0.00 0.00 0.20 0.20 0.00
59 org.mozilla.firefox 482 76.56 124.34 63.47 0.00
60 firefox 84496 76.70 89.18 10.58 5.59 124.55 63.48 0.00
62 This sample was taken while the following programs were running:
64 - Firefox Beta (single process, invoked from the Mac OS dock, shown in
65 the `org.mozilla.firefox` coalition.)
66 - Firefox Nightly (multi-process, invoked from the command line, shown
67 in the `com.apple.Terminal` coalition.)
71 The grouping of parent and child processes (in coalitions) is obvious.
72 The meaning of the columns is as follows.
74 - **Name**: Coalition/process name. Process names within coalitions
76 - **ID**: Coalition/process ID number.
77 - **CPU ms/s**: CPU time used by the coalition/process, per second,
78 during the sample period. The sum of the process values typically
79 exceeds the coalition value slightly, for unknown reasons.
80 - **User%**: Percentage of that CPU time spent in user space (as
81 opposed to kernel mode.)
82 - **Deadlines (\<2 ms, 2-5 ms)**: These two columns count how many
83 \"short\" timers woke up threads in the process, per second, during
84 the sample period. High frequency timers, which typically have short
85 time-to-deadlines, can cause high power consumption and should be
87 - **Wakeups (Intr, Pkg idle)**: These two columns count how many
88 wakeups occurred, per second, during the sample period. The first
89 column counts interrupt-level wakeups that resulted in a thread
90 being dispatched in the process. The second column counts \"package
91 idle exit\" wakeups, which wake up the entire package as opposed to
92 just a single core; such wakeups are particularly expensive, and
93 this count is a subset of the first column\'s count.
94 - **GPU ms/s**: GPU time used by the coalition/process, per second,
95 during the sample period.
99 - Smaller is better --- i.e. results in lower power consumption ---
100 for all of these measurements.
101 - There is some overlap between the two \"Deadlines\" columns and the
102 two \"Wakeups\" columns. For example, firing a single sub-2ms
103 deadline can also cause a package idle exit wakeup.
104 - Many of these measurements are also obtainable by passing the
105 `TASK_POWER_INFO` flag and a `task_power_info` struct to the
106 `task_info` function.
107 - By default, the coalitions/processes are sorted by a composite value
108 computed from several factors, though this can be changed via
109 command-line options.
111 ## Other measurements
113 `powermetrics` can also report measurements of backlight usage, network
114 activity, disk activity, interrupt distribution, device power states,
115 C-state residency, P-state residency, quality of service classes, and
116 thermal pressure. These are less likely to be useful for profiling
117 Firefox, however. Run with the `--show-all` to see all of these at once,
118 but note that you\'ll need a very wide window to see all the data.
120 Also note that `powermetrics -h` is a better guide to the the
121 command-line options than `man powermetrics`.
125 You can use the `mach power` command to run `powermetrics` in
126 combination with `rapl` in a way that gives the most useful summary
127 measurements for each of Firefox, Chrome and Safari. The following is
130 total W = _pkg_ (cores + _gpu_ + other) + _ram_ W
131 #01 17.14 W = 14.98 ( 5.50 + 1.19 + 8.29) + 2.16 W
133 1 sample taken over a period of 30.000 seconds
135 Name ID CPU ms/s User% Deadlines (<2 ms, 2-5 ms) Wakeups (Intr, Pkg idle) GPU ms/s
136 com.google.Chrome 500 439.64 585.35 218.62 19.17
137 Google Chrome Helper 67319 284.75 83.03 296.67 0.00 454.05 172.74 0.00
138 Google Chrome Helper 67304 55.23 64.83 0.03 0.00 9.43 4.33 19.17
139 Google Chrome 67301 63.77 68.09 29.46 0.13 76.11 22.26 0.00
140 Google Chrome Helper 67320 38.30 66.70 17.83 0.00 45.78 19.29 0.00
141 com.apple.WindowServer 68 102.58 112.36 43.15 80.52
142 WindowServer 141 103.03 58.19 60.48 6.40 112.36 43.15 80.53
143 com.apple.Safari 499 267.19 110.53 46.05 1.69
144 com.apple.WebKit.WebContent 67372 190.15 79.34 2.02 0.14 129.28 53.79 2.33
145 com.apple.WebKit.Networking 67292 65.23 52.74 0.07 0.00 4.33 1.40 0.00
146 Safari 67290 29.09 77.65 0.23 0.00 7.13 3.37 0.00
147 com.apple.Safari.SearchHelper 67371 13.88 91.18 0.00 0.00 0.36 0.05 0.00
148 com.apple.WebKit.WebContent 67297 0.81 56.84 0.10 0.00 2.20 1.30 0.00
149 com.apple.WebKit.WebContent 67293 0.46 76.40 0.03 0.00 0.57 0.20 0.00
150 com.apple.WebKit.WebContent 67295 0.24 67.72 0.00 0.00 0.90 0.37 0.00
151 com.apple.WebKit.WebContent 67298 0.17 59.88 0.00 0.00 0.50 0.13 0.00
152 com.apple.WebKit.WebContent 67296 0.07 43.51 0.00 0.00 0.10 0.03 0.00
153 kernel_coalition 1 111.76 724.80 213.09 0.12
154 kernel_task 0 107.06 0.00 5.86 0.00 724.46 212.99 0.12
155 org.mozilla.firefox 498 92.17 212.69 75.67 1.81
156 firefox 63865 61.00 87.18 1.00 0.87 25.79 9.00 1.81
157 plugin-container 67269 31.49 72.46 1.80 0.00 186.90 66.68 0.00
158 com.apple.WebKit.Plugin.64 67373 55.55 74.38 0.74 0.00 9.51 3.13 0.02
159 com.apple.Terminal 109 6.22 0.40 0.23 0.00
160 Terminal 208 6.25 92.99 0.00 0.00 0.33 0.20 0.00
162 The `rapl` output is first, then the `powermetrics` output. As well as
163 the browser processes, the `WindowServer` and kernel tasks are shown
164 because browsers often trigger significant load in them.
166 The default sample period is 30,000 milliseconds (30 seconds), but that
167 can be changed with the `-i` option.