M4 Pro MBP - 68W total power consumption, but only 40W processor power? What's consuming the remaining 28W?

[TommyYOyoyo]

While benchmarking my 14" M4 Pro 12c MacBook Pro on Cinebench 2024 multi-core, I noticed that the system is using a total of around 68W of power while the CPU power from Powermetrics/iStat/Mx Power Gadget was merely 40W. This means that there's an extra 28W consumed by other components... but what are they?

I monitored power consumption using Powermetrics, iStat Menus, Stats and Mx Power Gadget and noticed that the display was only consuming around 1-2W of power, the RAM around 5W and the fans around 3W. So that leaves us around 18W of remaining power consumed by other components...

I also tried while unplugged with Wi-Fi & Bluetooth disabled, but it's still drawing similar amount of power. My brightness has been set to 50%.

Model: 14" MacBook Pro M4 Pro 12-core CPU, 16-core GPU, 48GB RAM, 1TB SSD

Thanks in advance!

 

[okkibs]

These numbers are approximations and might not contain every component necessary to power the device. The best way to measure is to plug the charger into a power meter and make sure the battery is fully charged, then run the test and measure power draw at the wall. The chargers are very efficient so what you measure at the wall will reflect the actual power draw pretty well as long as the power meter shows correct numbers in the first place.

In any case, there is nothing wrong with the Mac, it draws as much power as it needs so beyond satisfying your curiosity there is nothing you can change about it.

 

[TommyYOyoyo]

These numbers are approximations and might not contain every component necessary to power the device. The best way to measure is to plug the charger into a power meter and make sure the battery is fully charged, then run the test and measure power draw at the wall. The chargers are very efficient so what you measure at the wall will reflect the actual power draw pretty well as long as the power meter shows correct numbers in the first place.

In any case, there is nothing wrong with the Mac, it draws as much power as it needs so beyond satisfying your curiosity there is nothing you can change about it.

Thanks for the response, I used AlDente to pause battery charging, and the wall wattage correlated fairly tightly with the "Total System Power" measurement of iStat and Stats (~68W), meaning that the ~28W extra power used by the remaining components other than CPU+GPU+ANE should be real.

Sorry, I can't quite help with my curiosity 😅😅😅

 

[JPack]

noticed that the display was only consuming around 1-2W of power,

That doesn't even pass a sanity check.

LCD backlights consume about 5-20W depending on brightness. The top two things that affect battery life are processor power and display brightness.

The SSD will be about 5W. Even the Nidec cooling fans will draw a couple watts each.

 

[TommyYOyoyo]

After reading the thread at forums.macrumors.com/threads/m3-max-14c-powermetrics-wattage-wrong-14-vs-16.2414321/ , I'm now extremely confused.

TLDR; the OP of that thread found out that an IOkit SMC sensor "PHPS" seemed to reflect more accurately the processor package power under load than Powermetrics (it was closer to total system consumption), although the sensor metrics were retrieved by 3rd party apps such as Stats.

While benchmarking on Cinebench 2024, Powermetrics showed that my CPU+GPU+ANE are consuming alltogether 40W, while the "PHPS" sensor was showing 53W of consumption. The system total power consumption during this was around 68W.

When running the script written by Leman in that same thread, Powermetrics showed that my CPU+GPU+ANE are consuming alltogether 48W, which was already higher than the M4 Pro TDP I found online, but "PHPS" sensor read an even higher 54W power consumption, while the system total consumption was around 58W.

Does this mean that the M4 family's TDPs published online (e.g. M4 Pro 12-core is known to have a 38W TDP) are not measured accurately, and that the true package TDP is much higher?

I don't know which sensor to trust anymore... Powermetrics or IOkit PHPS?

 

[JPack]

Does this mean that the M4 family's TDPs published online (e.g. M4 Pro 12-core is known to have a 38W TDP) are not measured accurately, and that the true package TDP is much higher?

That's another number that fails a sanity check. Apple's own M4 Pro documentation already ballparks the TDP at 80W.

140W - (15W *3 TB5) - (4.5W * 2 USB3) = 86W

Mac mini power consumption and thermal output (BTU) information - Apple Support

Learn about the power consumption and thermal output of Mac mini computers.

support.apple.com

This is substantiated by reviewer testing.

 

[TommyYOyoyo]

That's another number that fails a sanity check. Apple's own M4 Pro documentation already ballparks the TDP at 80W.

140W - (15W *3 TB5) - (4.5W * 2 USB3) = 86W

Mac mini power consumption and thermal output (BTU) information - Apple Support

Learn about the power consumption and thermal output of Mac mini computers.

support.apple.com

This is substantiated by reviewer testing.

Sorry if I wasn't clear, I meant the TDP of the M4 Pro SoC alone (CPU cores, GPU cores, cache, etc), not the entire system's Power consumption with all the components (SoC, SSD, display, etc).

The reason I'm confused is because, as the users from the other thread have pointed out, powermetrics' SoC power reading (40W) is substantially below the IOkit "PHPS" sensor reading (52W), which, according to them, was likely the SMC reading of the processor package power.

Since the SoC TDPs published online on https://en.wikipedia.org/wiki/Apple_M4 and websites like Notebookcheck or Nanoreview correlate only with powermetrics reading, while the PHPS sensor reads a much higher power consumption, I don't know which sensor to trust anymore.

Sorry for the confusion!