M4 Max EATS battery

[johnalan]

I just ran Cinebench's 10min GPU test on a fresh M4 Max top spec and it went from 100% battery to 82% in one 10min run.

Either this machine is a lemon or this is like Intel level battery use.

 

[Apple_Robert]

Why are you surprised? Fully taxing the machine with non-normal use for 10 minutes on battery is going to give that kind of results.

 

[johnalan]

Is that expected @Apple_Robert ? - not sure how else to test the battery, any ideas?

 

[Apple_Robert]

Is that expected @Apple_Robert ? - not sure how else to test the battery, any ideas?

Others here might disagree but, to me, your results are to be expected. If you want to test the battery, test it under normal use conditions (how you use the computer every day). Keep in mind that Apple‘s stated battery usage number on their website is done under lab conditions and not indicative of normal use. As such, M4 Mac users on MR and elsewhere can have different daily battery usage results.

 

[dmccloud]

I just ran Cinebench's 10min GPU test on a fresh M4 Max top spec and it went from 100% battery to 82% in one 10min run.

Either this machine is a lemon or this is like Intel level battery use.

1. That GPU test causes wattages to spike on any device, even Windows desktops, so I'm not surprised to see that type of battery drain.

2. If this is a new M4 Max machine, then the system is likely still in the process of calibrating battery life and charging curves, especially if you kept optimized battery charging enabled.

 

[iMacDragon]

Sounds within possible bounds for very high cpu/gpu usage, especially if screen is also very bright.

The long battery life figures assume light mostly idle usage at medium brightness.

 

[smithdr]

Nothing scientific, but I have been doing some work with my arrived today M4Max 16MBP/64GB/16 CPU/40 Core GPU using Davinci Resolve Studio. My impression is that uses more battery than a M1Max 16MBP/64GB/16 CPU/32 Core GPU for not much being that much faster, it gets hot and the fans come on quickly. If you have an M1Max 16MBP is the M4Max really worth the price for editing video?

 

[mi7chy]

I just ran Cinebench's 10min GPU test on a fresh M4 Max top spec and it went from 100% battery to 82% in one 10min run.

Either this machine is a lemon or this is like Intel level battery use.

Is that 16" or 14"? That's why you can't rely only on sponsored influencers since they don't discuss key things like battery life and fan noise under load. If it's draining 18% every 10 mins then battery will be drained just under 1 hour under load plus return of pre-2024 Intel fan noise and tethered to wall outlet.

 

[throAU]

Is that expected @Apple_Robert ? - not sure how else to test the battery, any ideas?

Do some maths...

The CPU/GPU can (and will, if you give it max workload with a benchmark) draw say 100 watts. They will be running flat out during the run, burning all the power they can to get the work done.

Running it for 1 hour will therefore draw ~100 watt hours of power. 10 minutes should draw approximately 1/6th of that. Or say, 15 watt/hours.

The 14" has something like a 72 watt/hr battery.

So napkin math - 20% for a run is not unexpected. Doing hard core max effort processing like that isn't something you do on battery. You plug into the wall for that sort of thing. This isn't exclusive to apple, intel or anyone else. Its basic physics.

However, unlike intel - less severe workloads you can run on battery, etc.

The difference between this and intel isn't (so much) max power draw during max effort - its the fact that doing "normal" workloads the CPU/GPU is going to be much more idle. Or, it will get more processing done in less time (and still burn less battery than intel for given amount of work completed). And if you do need max power for sustained workload - as with any laptop, plug it in.

 

[throAU]

For comparison:

An intel notebook with mobile Nvidia 4080 or 4090 will be burning through 3x that power and this is why they throttle a HEAP on battery. Because the power just isn't there to drive those intel furnaces and high end Nvidia GPUs on battery effectively.


You can do the rough math with any laptop - get its battery capacity in watt/hours, work out the max CPU/GPU power draw and that's how many watts/hours per hour it will burn at max load (approximately).

The important thing is how much work they can get done without running at max load on battery - basically as cpu/gpu load (and therefore clock frequency) rises, power draw goes up exponentially.

Which is why intel is so hot/power hungry these days - they're pushing frequency as hard as they can to keep up (rather than make the processor design more efficient in terms of performance per clock cycle) and it doesn't scale very well with power consumption.

 

[johnalan]

yep I then tested with 4k HDR video at medium brightness and it ran for an hour using 5% battery, which is good.

 

[pugxiwawa]

Is that 16" or 14"? That's why you can't rely only on sponsored influencers since they don't discuss key things like battery life and fan noise under load. If it's draining 18% every 10 mins then battery will be drained just under 1 hour under load plus return of pre-2024 Intel fan noise and tethered to wall outlet.

holy cow this sounds terrible.

 

[ProQuiz]

Is that 16" or 14"? That's why you can't rely only on sponsored influencers since they don't discuss key things like battery life and fan noise under load. If it's draining 18% every 10 mins then battery will be drained just under 1 hour under load plus return of pre-2024 Intel fan noise and tethered to wall outlet.

The guy in the video stated he was using High Power Mode. Wasn't it mentioned that High Power Mode significantly increases fan speed without doing anything else?

 

[Macintosh IIcx]

The guy in the video stated he was using High Power Mode. Wasn't it mentioned that High Power Mode significantly increases fan speed without doing anything else?

It will ramp up the fans substantially, but the M4 might be the first chip that can increase clock speed quite a bit if it is given more thermal headroom, at least the feature didn’t bring much to the M1 generation if I recall correctly. I’m not sure if anybody have made a controlled test of how much the performance increases with High Power Mode on the M4 Max, the 14” is surely not the right to use in that scenario.

 

[UnifiedMelody]

No surprises. Even testing my friends base 14" M4 Pro to render 1:1 500 images in LR only, the fans ramped up high && I saw the battery drop from 100% to 82% in 15-16 minutes.

Also M1 was very conservative, M4 is pushing the envelope which is why I'm not surprised the unit will get hot a bit quicker. The clock speed has gone up quite a fair bit from M1 to M4 [3.2 to 4.5]

 

[Grilled Cheese]

Nothing scientific, but I have been doing some work with my arrived today M4Max 16MBP/64GB/16 CPU/40 Core GPU using Davinci Resolve Studio. My impression is that uses more battery than a M1Max 16MBP/64GB/16 CPU/32 Core GPU for not much being that much faster, it gets hot and the fans come on quickly. If you have an M1Max 16MBP is the M4Max really worth the price for editing video?

Sorry to hear that. So you’re not seeing much performance increase from M1 Max to M4 Max with video editing? Export times much faster?

 

[arcite]

Nothing scientific, but I have been doing some work with my arrived today M4Max 16MBP/64GB/16 CPU/40 Core GPU using Davinci Resolve Studio. My impression is that uses more battery than a M1Max 16MBP/64GB/16 CPU/32 Core GPU for not much being that much faster, it gets hot and the fans come on quickly. If you have an M1Max 16MBP is the M4Max really worth the price for editing video?

I’ve been noticing this with other reviews. This is the price for Apple increasing the wattage. There is a price to be paid….that price is more fan noise, more heat, ….less battery. It’ll be interesting to see how these chips fare after several years of hard use.

 

[Chancha]

Max Tech's latest video demonstrate power consumption pretty well, particular in 3DMark the GPU alone takes 84W, and Cinebench the total package can add up to 110W. This is not yet counting the display, WiFi, attached bus powered device like external drives.

 

[johnalan]

Max Tech's latest video demonstrate power consumption pretty well, particular in 3DMark the GPU alone takes 84W, and Cinebench the total package can add up to 110W. This is not yet counting the display, WiFi, attached bus powered device like external drives.

so sounds like my machine is operating normally, given the battery is only 100 odd watt hours.

 

[throAU]

So in case anyone is curious.
I ran my m4 max 14” on battery last night for 6 hours doing basic web surfing/youtube with teams and email running.

In 5-6 hours on “automatic” power level battery dropped to 58%

 

[Varoudis]

like my 2018 i9
I you work with ML models or CPU multithreaded code that needs to run for 20mins you need a battery even if you are plugged into the wall!!

 

[smirking]

So in case anyone is curious.
I ran my m4 max 14” on battery last night for 6 hours doing basic web surfing/youtube with teams and email running.

In 5-6 hours on “automatic” power level battery dropped to 58%

I'm about to make a similar transition to one you made recently. Even similar specs! I'm on a 16" M1 Pro 16GB and expect to go to a 14" M4 Pro 48GB before the end of the year. Battery life is really important to me. After 3 years on an average day, I'm still getting 6-8 hours on battery on my M1 Pro doing general computing, light photo editing, running dev tools most of the day, and having 1-2 small local development app servers running in the background.

Are you seeing anything that would suggest that I shouldn't expect to get similar battery life moving up to an M4 Pro?

 

[throAU]

I'm about to make a similar transition to one you made recently. Even similar specs! I'm on a 16" M1 Pro 16GB and expect to go to a 14" M4 Pro 48GB before the end of the year. Battery life is really important to me. After 3 years on an average day, I'm still getting 6-8 hours on battery on my M1 Pro doing general computing, light photo editing, running dev tools most of the day, and having 1-2 small local development app servers running in the background.

Are you seeing anything that would suggest that I shouldn't expect to get similar battery life moving up to an M4 Pro?

Well, I've got mostly the same stuff I had installed on the M1 Pro by the time I migrated off it. Battery life seems basically the same (if not better) going from M1 Pro to m4 max both in 14". Under "normal" working conditions. i.e., no background renders, just mail, a few terminal apps, web pages, etc.

I'm a network admin, being a web developer I suspect a lot of what you do is similar laptop workload to me - a bunch of browsers, terminal sessions, Remote Desktop sessions, YouTube/video playback, emails, background file transfers, etc.

I think if you're going pro to pro you'll be impressed with the battery life gain if anything. I went pro to max and am happy. I don't run heavy workloads on battery though, anything I do that's "heavy" is when I'm at a desk (gaming, playing with blender, virtual machines, etc.).

Obviously, running the machine flat out the m4 series can burn through more power, but most people simply aren't doing that.

 

[chrono1081]

I just ran Cinebench's 10min GPU test on a fresh M4 Max top spec and it went from 100% battery to 82% in one 10min run.

Either this machine is a lemon or this is like Intel level battery use.

This is why I hate benchmarks, they're silly. You ran a 10 minute GPU test which is designed to max out the GPU, this is 100% to be expected.

You're not going to see that kind of drain in normal use but you're literally pushing the system to its limit with those kinds of tests.

 

[UnifiedMelody]

So in case anyone is curious.
I ran my m4 max 14” on battery last night for 6 hours doing basic web surfing/youtube with teams and email running.

In 5-6 hours on “automatic” power level battery dropped to 58%

Yeah I felt high power mode didn't really do much for exporting and stuff like that so now I just would leave it on automatic [or low power mode if im desperate to inch battery life out , which in the future case never] Done my own personal benching [nothing like their cinebench lol] with some real life workflow for video and photo... barely notice diff