Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

8 core vs 10 core battery life?

BluAffiliate said:
Is it safe to assume that the 8 core 14' has better battery life than the 10 core?
Click to expand...

No. Most probably there is zero difference. The lottery of slightly better or worse batteries is probably a bigger effect.

If you happen to do work which uses all available cores then the 10 core will use more power, but finish the job quicker. If you use all cores until the battery dies then yes the 8 will last longer but will have done no more work during its extra runtime so what’s the point.
 
  • Like
Reactions: amt2002, T'hain Esh Kelch and BluAffiliate
Nope,both use the same amount of power ,and probably heat as much
Even though 1core per cluster has been disabled ,it seems that the clusters still received the same power as 10c version

Edit : since they use as much power,but aren't as powerful ,it means the runtime is actually longer ,hence the 8c most likely has a WORSE battery than the 10c.crazy ****

Notebookcheck review :

"There are AMD and Intel CPUs that offer much more performance, but they also need more threads for the same performance and Apple's SoC is much more efficient. The full package power is just 7 W (4 W for the CPU core) in single-core tests and just 24.7 W (with ~21 W for the CPU cores) in multi-tests. However, it looks like the CPU clusters consume the same amount of power as in the M1 Pro with 10 cores, which offers more performance."
 
Can we verify this Notebookcheck statement?

It would be a very strange engineering decision to feed the same amount of power to an 8 core and a 10 core, where the cores are running at the same frequency… makes no sense.

Does anyone know how?
 
I quote a friend of mine here
"
if it's still 2 clusters of 4 cores with 2 cores disabled.

it's not like you can cut the power to those cores

doesn't work like that

also those 8-core are possibly lower-binned version

so it may require higher voltage

for stable operation

thereby higher power consumption

you still need to feed power into the cluster

so either the remaining 3 cores, for example, consume higher power

or the disabled core still gets that power

even though disabled.

because it's not independent cores but grouped as a cluster

so the whole cluster will get feed with the same voltage

the rest idk

for example in SD SoC

the prime core have the same voltage as the Middle ones

so if prime requires high voltage (high freq) then the middle will get same too

which will be wasted

"
 
0089294 said:
I quote a friend of mine here
"
if it's still 2 clusters of 4 cores with 2 cores disabled.

it's not like you can cut the power to those cores

doesn't work like that

also those 8-core are possibly lower-binned version

so it may require higher voltage

for stable operation

thereby higher power consumption

you still need to feed power into the cluster

so either the remaining 3 cores, for example, consume higher power

or the disabled core still gets that power

even though disabled.

because it's not independent cores but grouped as a cluster

so the whole cluster will get feed with the same voltage

the rest idk

for example in SD SoC

the prime core have the same voltage as the Middle ones

so if prime requires high voltage (high freq) then the middle will get same too

which will be wasted

"
Click to expand...

Interesting… so what happens to the power fed to a disabled core? Is it just turned into heat?

The disabled core, even if in a cluster, should be using next to no power. If that’s the case then it would mean the other 3 cores are drawing more power than “normal” and should have higher performance. However, that’s not being seen in benchmarks.
 
Last edited:
It will be negligible. Where people are noticing a difference is when you start to increase the GPU cores.
 
bluegt said:
Interesting… so what happens to the power fed to a disabled core? Is it just turned into heat?

The disabled core, even if in a cluster, should be using next to no power. If that’s the case then it would mean the other 3 cores are drawing more power than “normal” and should have higher performance. However, that’s not being seen in benchmarks.
Click to expand...
Ikr ? Energy doesn't disappear like that,it has to transform.im guessing it'll heat up.not saying it'll be hotter than the 10c but likely as hot .

I'm not knowledgeable enough to answer you my man..
However what It seems is that they kept the same power per cluster.you won't get more Performance because of that,you see,it depends on the cores clock ,and they are identical.its either Enough power or too much.and it's too much lol

1)its not easy to drain energy per core apparently
2)it could be that apple didn't want the binned model ,being cheaper,to outperform the 10c ,and be cooler and more efficient.i could see apple bastards doing it tbf
 
0089294 said:
Ikr ? Energy doesn't disappear like that,it has to transform.im guessing it'll heat up.not saying it'll be hotter than the 10c but likely as hot .

I'm not knowledgeable enough to answer you my man..
However what It seems is that they kept the same power per cluster.you won't get more Performance because of that,you see,it depends on the cores clock ,and they are identical.its either Enough power or too much.and it's too much lol

1)its not easy to drain energy per core apparently
2)it could be that apple didn't want the binned model ,being cheaper,to outperform the 10c ,and be cooler and more efficient.i could see apple bastards doing it tbf
Click to expand...

Hahaha, I am tempted to upgrade my 8 core to 10 core now. Apple is very good at convincing me to spend money.
 
Has anyone seen performance or battery benchmarks between the
14" 8 CPU / 14 GPU
14" 10 CPU / 14 GPU

If in fact the CPU power draw is the same between the 8 core and 10 core, then presumably the battery life between these models should be the same.
 
I saw a dude on youtube or wherever edit 8k video on base 14 inch, and it ran awesome. So I wonder if it really matters in realtime work.
 
Last edited:
  • Like
Reactions: Queen6
There is a relatively minor difference. The 8 core will net the best battery life under load. For general productivity I don't think you will notice much of a difference. This guy does the tests on the following configs in the 14":

CPU/GPU
8/14
10/16
10/24

 
  • Like
Reactions: haruhiko
nylon said:
There is a relatively minor difference. The 8 core will net the best battery life under load. For general productivity I don't think you will notice much of a difference. This guy does the tests on the following configs in the 14":

CPU/GPU
8/14
10/16
10/24

Click to expand...
yup i just saw this video.however im wondering if we will ever see how it behaves under light load ,which is what really interstets me when im traveling
 
bluegt said:
Has anyone seen performance or battery benchmarks between the
14" 8 CPU / 14 GPU
14" 10 CPU / 14 GPU

If in fact the CPU power draw is the same between the 8 core and 10 core, then presumably the battery life between these models should be the same.
Click to expand...
I think between those two, the results and tests I’ve seen are just what one expects. In CPU intensive tasks, its about 20% faster. I went with it since I thought that sort of slight time save, and NOT having a binned CPU, so a clear quality 10 core, I figured well that might be a little insurance policy as well. I’m sure apple knows pretty well how to figure out which CPU’s have some problem cores, and turn them off and make an 8 core CPU, but I figured lets just go with the solid 10 core tested version.
 
  • Like
Reactions: haruhiko
0089294 said:
I quote a friend of mine here
"
if it's still 2 clusters of 4 cores with 2 cores disabled.
it's not like you can cut the power to those cores"
Click to expand...

I work in SoC chip design. The power being used by the 2 disabled cores really depends on how the power delivery has been architected. If each core has it's own switched power domain (seems doubtful) then a disabled core can be completely powered down and would use no power. If all the cores are on the same supply (most likely) then the core can be disabled by shutting down the clocks inside of it. That would reduce dynamic power to virtually zero, but not change the inherent leakage power of the CMOS gates. And leakage power at small nodes is a big chunk of the overall power.
 
  • Like
Reactions: amt2002, Sanpete and T'hain Esh Kelch
jasonp99 said:
I work in SoC chip design. The power being used by the 2 disabled cores really depends on how the power delivery has been architected. If each core has it's own switched power domain (seems doubtful) then a disabled core can be completely powered down and would use no power. If all the cores are on the same supply (most likely) then the core can be disabled by shutting down the clocks inside of it. That would reduce dynamic power to virtually zero, but not change the inherent leakage power of the CMOS gates. And leakage power at small nodes is a big chunk of the overall power.
Click to expand...
Thanks for the insight man !
What exactly is dynamic power ? And on average how much does leakage power represents compared to the 10c fly enabled variant would u say ?
You said it's a big chunk ,but I'm curious to know what you mean by that.somethinf like 50% of the power,or closer to 90% lol
 
Datapoint for the 8-Core Base 14":

Restarted
Background Apps: Windscribe (VPN), BitDefender and Google Drive
Idled 5 Minutes, Brightness at 6 Ticks
Total Power Consumption 0.39Wh
Average Wattage: 4.7W

Powermetrics indicates CPU power consumption is: 78mW
E-Cluster: 8mW
P0-Cluster: 65mW
P1-Cluster: 6mW

So whether 8 or 10 core, if the cores are idle, power consumption is next to nil.
 
  • Like
Reactions: Sanpete
Question remains on full power mode. Can someone with a 10-core CPU please help to test.

Steps:
1. Open Terminal, type sudo powermetrics
2. Go to https://cpux.net/cpu-stress-test-online
3. Set to 32 Threads, 100% Power, Click Start
4. Watch powermetrics for 2-3 minutes, and report on the max results
*note, keep window focus on the browser

For my 8-Core Base 14" I maxed out at:
CPU Power: ~27300mW
Package Power: ~32600mW
 
  • Like
Reactions: Love-hate 🍏 relationship
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back