Will Two Extra GPU Cores or More RAM Significantly Shorten Battery in M2 Air?

[MuckrakerJG]

If I max out the M2 Air by taking the 10 core GPU model to 24 GB Ram, will that be a big drop in battery life compared to 8 cores and 8 RAM?

I see Apple's battery testing fine print on the tech specs mention they used the base 1199 model.

I also remember when MacBook Pro came out The Verge and others confirmed that more GPU cores suck up extra battery even when not in use for an intensive task. So, 16" M1 Pro for $2499 has better battery life during basic word processing and web browsing than 16" M1 MAX for $3499 because there are fewer cores there to turn on or power, even if the task isn't using them at the moment.

And yes, I know a 14" MBP might be better value, but I like the Air specifically because it is more lightweight and fan-less. I use my current M1 Air in bed on top of the covers a lot and that's not something you are supposed to do with a laptop with a fan.

 

[DrCC]

Once the new devices ship, reviewers will most likely test this out. Until then, it is just speculation. 50-50 change of shortening the battery life or not.

 

[UBS28]

If you stress the 10 cores, ofcourse it will drain the battery more.

The M1 Max under maximum stress also has a bigger power draw than the M1 Pro.

 

[Gelam]

I am in the same boat with the OP.

I am curious of the power consumption of 8 vs 10 GPU cores when:
1. At idle.
2. Under light tasks like word, excel, powerpoint, web browsing, videos, light pdf viewing, light photo editing.
3. Under medium loads like 100 safari tabs, big pdf or multiple pdf opened, big excel, long word documents, big powerpoint presentation.

Is this all that you have in mind as well OP?

 

[Kazgarth]

Don't think so.

The M1 Pro & M1 Max had a difference of 16 GPU cores (that's 2x MBA total GPU count) and the effect on the battery life was only 1-2 hours under light load.

This M2 variants have only 2 GPU core difference, so the difference will be very negligible on the battery life.

 

[Gelam]

I hope it will be like your case. Would love to have the maximum battery life, but also is temped to get 10gpu option to see how RE:Village runs on a fanless ultrabook with max settings .

 

[Gelam]

Don't think so.

The M1 Pro & M1 Max had a difference of 16 GPU cores (that's 2x MBA total GPU count) and the effect on the battery life was only 1-2 hours under light load.

This M2 variants have only 2 GPU core difference, so the difference will be very negligible on the battery life.

I hope it will be like your case. Would love to have the maximum battery life, but also is temped to get 10gpu option to see how RE:Village runs on a fanless ultrabook with max settings .

 

[T'hain Esh Kelch]

Likely the display and the rest of the electronics will do most of the power load, so the effect should be relatively minimal. That being said, Apple says the new model has 15h/18h battery life, same as the previous one, but that is with a 5% larger battery. So it is safe to say it uses slightly more power, but not a lot. If you do something that lets the CPU run continuously, say video encoding or similar, then you will likely notice a few hours less battery, otherwise, not likely to see a difference. But your jobs will also finish slightly faster, and chances are you won't be doing 12h video encodes sitting on the train, without a power plug.

Personally I would not pick the M1 over the M2, if battery life was the only parameter I cared about.

 

[leman]

I doubt you will see any significant difference on idle/light load, as the extra energy consumed by the GPU cores is likely to be balanced out by higher performance (less time needed to perform the work). There will likely be a difference when doing heavy GPU work but then the battery life hardly matters anyway.

RAM has very little impact on battery life in Apple Silicon because it is run at ridiculously low voltages and consumes very little power under normal operation.

 

[Gelam]

Likely the display and the rest of the electronics will do most of the power load, so the effect should be relatively minimal. That being said, Apple says the new model has 15h/18h battery life, same as the previous one, but that is with a 5% larger battery. So it is safe to say it uses slightly more power, but not a lot. If you do something that lets the CPU run continuously, say video encoding or similar, then you will likely notice a few hours less battery, otherwise, not likely to see a difference. But your jobs will also finish slightly faster, and chances are you won't be doing 12h video encodes sitting on the train, without a power plug.

Personally I would not pick the M1 over the M2, if battery life was the only parameter I cared about.

I agree I would not go M1 for sure. However, my workload is a constant light/medium in that I have 100+ safari tabs and multiple large document open at once that I do not close for months, if not the entire life of the machine. So it is constant not a temporal workload like rendering a video.

This information is to see the GPU power sip at these constant load and see if there will be differences in battery life.
I think my case is a little niche though so I understand your view point.

 

[leman]

I agree I would not go M1 for sure. However, my workload is a constant light/medium in that I have 100+ safari tabs and multiple large document open at once that I do not close for months, if not the entire life of the machine. So it is constant not a temporal workload like rendering a video.

This information is to see the GPU power sip at these constant load and see if there will be differences in battery life.
I think my case is a little niche though so I understand your view point.

That is not a constant load. You can have a million Safari tabs open, and your machine will still essentially be running idle. Unless an invisible tab is doing some heavy background work like cryptomining (but then it will be throttled to oblivion by all major browsers), it is not consuming any CPU or GPU resources.

 

[T'hain Esh Kelch]

I agree I would not go M1 for sure. However, my workload is a constant light/medium in that I have 100+ safari tabs and multiple large document open at once that I do not close for months, if not the entire life of the machine. So it is constant not a temporal workload like rendering a video.

This information is to see the GPU power sip at these constant load and see if there will be differences in battery life.
I think my case is a little niche though so I understand your view point.

That is not a constant load. RAM may be reserved, but your SOC pretty much sits idle for all of that. Check your Activity Viewer.

 

[jav6454]

More memory won't affect power draw as much as an extra 2 GPU cores.

 

[Ponylover52]

No using the specced out config day to day wont use more battery, and if you look at the A15 architecture review, the new core architecture is actually more efficient under load compared to the A14. I wouldn’t be shocked to see users beating the claimed run times by Apple. Obviously hammering the chip will vary this but under normal day to day use don’t be shocked to get excellent battery life. My M1 Air is continuously giving me 11 hours use over the 10 quoted.

The Apple A15 SoC Performance Review: Faster & More Efficient

www.anandtech.com

 

[Gelam]

That is not a constant load. You can have a million Safari tabs open, and your machine will still essentially be running idle. Unless an invisible tab is doing some heavy background work like cryptomining (but then it will be throttled to oblivion by all major browsers), it is not consuming any CPU or GPU resources.

That is not a constant load. RAM may be reserved, but your SOC pretty much sits idle for all of that. Check your Activity Viewer.

I see, that is new knowledge. I would have thought that more background apps/tabs might cause the processor a little more power to maintain and swap memories.

More memory won't affect power draw as much as an extra 2 GPU cores.

Here are the information I have gathered so far on the differences in battery life for the MacBook Air M1 with 7 vs 8 core GPU.

7 Core GPU

Apple MacBook Air 2020 M1 Entry Review: Apple M1 CPU humbles Intel and AMD

The M1 based entry-level Macbook in our extensive technical test.

www.notebookcheck.net

8 Core GPU

Apple MacBook Air 2020 Review: Should you get the more powerful version of the M1 processor?

Notebookcheck reviews the new Apple MacBook Air 2020 M1. Do you really need the additional GPU core?

www.notebookcheck.net

According to the test the 7 core GPU has a longer battery life by 28 minutes on "Wifi Websurfing", but shorter battery life by 16 minutes on "Wifi Surfing max. Brightness". With this conflicting data, I cannot conclude anything. Keep in mind that this does not include "Low Power Mode" and the results from 8 core GPU "Wifi Websurfing" might be a fullscreen test like the "Wifi Websurfing (Fullscreen)" test on 7 core GPU. Some times data from these sites tend to miss comparable information or is not standardized.

No using the specced out config day to day wont use more battery, and if you look at the A15 architecture review, the new core architecture is actually more efficient under load compared to the A14. I wouldn’t be shocked to see users beating the claimed run times by Apple. Obviously hammering the chip will vary this but under normal day to day use don’t be shocked to get excellent battery life. My M1 Air is continuously giving me 11 hours use over the 10 quoted.

The Apple A15 SoC Performance Review: Faster & More Efficient

www.anandtech.com

That is very nice to see your MacBook air running 1 hour than quoted. However, was it not 15 hour web surfing and 18 hour movie playback that was quoted?

Also here is the data I have found so far for M1 vs M2 MacBook Pro.

M1 MacBook Pro

Apple MacBook Pro 13 2020 Laptop Review: The entry-level Pro also gets the M1 performance boost

Notebookcheck reviews the entry-level MacBook Pro 13 2020 with Apple's new M1 processor. Does the active cooling solution make a noticeable difference?

www.notebookcheck.net

M2 MacBook Pro

Apple MacBook Pro 13 2022 M2 Laptop Review – Debut for the new Apple M2

Notebookcheck reviews the updates 2022 Apple MacBook Pro 13 with the brand-new Apple M2 SoC and 16 GB RAM.

www.notebookcheck.net

Both tests were at 150 nits display brightness.
According to the results, the evidence points out that the M2 chip drains more energy at least at idle/low workloads. Or it could be that the new safari version 15.5 of this year runs more efficiently, since the video playback test resulted in the same battery life. Here is another test by Dave2D that contradicts with the test above.

This results showed the M2 being more efficient at light load.
It should be noted that this does not indicate if the M1 results were originally from a year ago or its from recently where the battery would have already degraded a little.

So here you can see why I cannot conclude anything because the information out there is so incomplete and contradicts each other.

 

[Gelam]

Also there are rumors that the M2 Pro will have a 3nm process instead of the 5nm on M2. It came from some reliable leakers, however, it I am skeptical and hope that is not the case as it would alienate the purchase of M2 air so much

 

[Gelam]

If I max out the M2 Air by taking the 10 core GPU model to 24 GB Ram, will that be a big drop in battery life compared to 8 cores and 8 RAM?

I see Apple's battery testing fine print on the tech specs mention they used the base 1199 model.

I also remember when MacBook Pro came out The Verge and others confirmed that more GPU cores suck up extra battery even when not in use for an intensive task. So, 16" M1 Pro for $2499 has better battery life during basic word processing and web browsing than 16" M1 MAX for $3499 because there are fewer cores there to turn on or power, even if the task isn't using them at the moment.

And yes, I know a 14" MBP might be better value, but I like the Air specifically because it is more lightweight and fan-less. I use my current M1 Air in bed on top of the covers a lot and that's not something you are supposed to do with a laptop with a fan.

OP, this thread have done some power draw testing for M1 Pro and M1 Max MacBook Pros.
This might help you extrapolate how the power draw will be like between the M2 8 and 10 GPU cores.

https://forums.macrumors.com/threads/request-can-some-with-m1-max-run-this-powermetrics-code-to-test-idle-load-power-draw.2320968/

 

[leman]

I see, that is new knowledge. I would have thought that more background apps/tabs might cause the processor a little more power to maintain and swap memories.

Backgrounds tabs usually either don't do any work — or if they do (e.g. some background processing), they will be heavily throttled by the browser/system. The way how this works can be counter-intuitive, as what really drains energy is the switching to the high-power state. This is why modern systems will try to group background processing requests and complete them in a single window of high activity. Of course, the efficiency cores on the Apple Silicon change the story significantly again. They consume so little power that their contribution becomes pretty much neglectable.

So here you can see why I cannot conclude anything because the information out there is so incomplete and contradicts each other.

Yeah, there is a trend, but it's still pretty much within the margin of error. Of course, the battery life you will see in your workflow will be very different. I think it's safe to assume that the practical differences are so small that have no noteworthy impact. Half and hour here or there is not something you can qualify reliable when it comes to computer operation.

 

[jav6454]

[...]

Here are the information I have gathered so far on the differences in battery life for the MacBook Air M1 with 7 vs 8 core GPU.

7 Core GPU

Apple MacBook Air 2020 M1 Entry Review: Apple M1 CPU humbles Intel and AMD

The M1 based entry-level Macbook in our extensive technical test.

www.notebookcheck.net

View attachment 2024765
8 Core GPU

Apple MacBook Air 2020 Review: Should you get the more powerful version of the M1 processor?

Notebookcheck reviews the new Apple MacBook Air 2020 M1. Do you really need the additional GPU core?

www.notebookcheck.net

View attachment 2024766 According to the test the 7 core GPU has a longer battery life by 28 minutes on "Wifi Websurfing", but shorter battery life by 16 minutes on "Wifi Surfing max. Brightness". With this conflicting data, I cannot conclude anything. Keep in mind that this does not include "Low Power Mode" and the results from 8 core GPU "Wifi Websurfing" might be a fullscreen test like the "Wifi Websurfing (Fullscreen)" test on 7 core GPU. Some times data from these sites tend to miss comparable information or is not standardized.



That is very nice to see your MacBook air running 1 hour than quoted. However, was it not 15 hour web surfing and 18 hour movie playback that was quoted?

Also here is the data I have found so far for M1 vs M2 MacBook Pro.

M1 MacBook Pro
View attachment 2024768

Apple MacBook Pro 13 2020 Laptop Review: The entry-level Pro also gets the M1 performance boost

Notebookcheck reviews the entry-level MacBook Pro 13 2020 with Apple's new M1 processor. Does the active cooling solution make a noticeable difference?

www.notebookcheck.net

M2 MacBook Pro
View attachment 2024769

Apple MacBook Pro 13 2022 M2 Laptop Review – Debut for the new Apple M2

Notebookcheck reviews the updates 2022 Apple MacBook Pro 13 with the brand-new Apple M2 SoC and 16 GB RAM.

www.notebookcheck.net

Both tests were at 150 nits display brightness.
According to the results, the evidence points out that the M2 chip drains more energy at least at idle/low workloads. Or it could be that the new safari version 15.5 of this year runs more efficiently, since the video playback test resulted in the same battery life. Here is another test by Dave2D that contradicts with the test above.
View attachment 2024777

This results showed the M2 being more efficient at light load.
It should be noted that this does not indicate if the M1 results were originally from a year ago or its from recently where the battery would have already degraded a little.

So here you can see why I cannot conclude anything because the information out there is so incomplete and contradicts each other.

1 core won't make a difference.