Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
It'll only take power from one port. This was mentioned at the keynote
Ah. Don't recall hearing that, but makes sense.

I wonder if a 3rd party 100w power supply would provide any additional juice or if the MBP is designed for a maximum of 87w of power delivery.
 
Is this issue widespread? I owned a 2015 with the m370x for a time and I didn't have this problem. If this is common on the 2016's I think I'll definitely cancel my order...
 
Dell and asus have been guilty of this all you have to do is search. It's quite common

That is not a specific example. I'm looking for a specific model which can't stay running with its provided power supply.

I note that I saw this happen once with a Toshiba... But it was using a third-party power supply that wasn't even close to the machine's listed rating. Using the actual provided power supply? Works fine.

My ASUS gaming laptops have all had adequate power supplies. Sometimes they were large bricks, but they were able to provide the power needed. I have a recentish ASUS and a Dell, and they both provide 180W power supplies, and they appear to run just fine under load.
 
That is not a specific example. I'm looking for a specific model which can't stay running with its provided power supply.

I note that I saw this happen once with a Toshiba... But it was using a third-party power supply that wasn't even close to the machine's listed rating. Using the actual provided power supply? Works fine.

My ASUS gaming laptops have all had adequate power supplies. Sometimes they were large bricks, but they were able to provide the power needed. I have a recentish ASUS and a Dell, and they both provide 180W power supplies, and they appear to run just fine under load.
A specific example off the top of my head would be the acer predator 17 i7 980mx. When under heavy load it can drain the battery. When the battery gets to 30% it stops using it and underclocks both the cpu and gpu.
 
That's not the solution, though. I'm using gaming as an example because that's where I'm seeing it right now. I do production work throughout the year that often requires an overnight render because it's effects-heavy or extremely high res for floor-to-ceiling displays or something of that nature. If I start one of these renders and it takes longer than six hours, the computer will flat out fail at the task. That, to me, is absolutely crazy. Gaming is not the only lengthy, high load situation that the owner of a quad core+dGPU notebook is going to encounter.

It happens all the time. I do transcode work for the company on my 12-core Hackintosh that can take up to two full days of processing to complete. A weekend-long shoot that generates 4TB of 6K footage which needs to be converted down to 1080p proxies for smooth post work - if I'm on the road, this limitation means I can't do it.

I understand and this means that the MBP is not good for your works.
 
A specific example off the top of my head would be the acer predator 17 i7 980mx. When under heavy load it can drain the battery. When the battery gets to 30% it stops using it and underclocks both the cpu and gpu.

Huh, right you are:

https://community.acer.com/t5/Preda...ery-whilst-gaming-on-power/m-p/407860#U407860

Interestingly, I note that general user consensus was the same there -- this is not acceptable behavior. And they mention a similar issue with a Dell machine, but Dell reacted by sending out improved power adapters.

I would tend to regard it as a pretty serious flaw, but you're right, it's not unique to Apple. I'm not sure how common it is. (Quick back-of-the-envelope math suggests that my current PC laptops have enough power supply for their likely maximum load, but I'm not totally sure.)

EDIT: Did some science. I was able to get my laptop loaded enough that it appeared to be gradually losing charge, but then it regained charge. I assume must have done some kind of throttling, but I didn't see anything. Maybe a background task finished.
 
Last edited:
Huh, right you are:

https://community.acer.com/t5/Preda...ery-whilst-gaming-on-power/m-p/407860#U407860

Interestingly, I note that general user consensus was the same there -- this is not acceptable behavior. And they mention a similar issue with a Dell machine, but Dell reacted by sending out improved power adapters.

I would tend to regard it as a pretty serious flaw, but you're right, it's not unique to Apple. I'm not sure how common it is. (Quick back-of-the-envelope math suggests that my current PC laptops have enough power supply for their likely maximum load, but I'm not totally sure.)

EDIT: Did some science. I was able to get my laptop loaded enough that it appeared to be gradually losing charge, but then it regained charge. I assume must have done some kind of throttling, but I didn't see anything. Maybe a background task finished.
It's basically penny pinching that companies will justify with the portability speech unfortunately.
 
On further reflection, I think there's a really significant distinction:

The Acer users were complaining about the machine slowing down. But you'll notice, it slowed down. It didn't shut down.

The users who have the huge workloads can live with a machine which slows down a bit on really long workloads, but still finishes them.

They can't get their work done on a machine which kills itself overnight.
 
  • Like
Reactions: eulslix
to be fair.... if you are maxing the CPU + GPU for more than 6 hours at a time, you should probably purchase a Desktop that can do this. Even if you need portability you can still use your desktop as a server and let it do all the heavy work while you access the results through the network. You are using the wrong tool for the job. I understand you probably want one device to suit all your needs, but there are sacrifices for your need for portability.

Desktops such as the iMac are great for doing heavy watt usage for days at a time.
 
On further reflection, I think there's a really significant distinction:

The Acer users were complaining about the machine slowing down. But you'll notice, it slowed down. It didn't shut down.

The users who have the huge workloads can live with a machine which slows down a bit on really long workloads, but still finishes them.

They can't get their work done on a machine which kills itself overnight.
Completely agree and tbh was never the behaviour on older macs. You could run them without a battery they just wouldn't run at full speed.
 
  • Like
Reactions: eulslix
Completely agree and tbh was never the behaviour on older macs. You could run them without a battery they just wouldn't run at full speed.

To be clear, though, I have not done the full 6 hour stress to see if it will shut down in OS X or simply under clock to hit the 87w max on AC. I'd hope it just downclocks.
 
That's not the solution, though. I'm using gaming as an example because that's where I'm seeing it right now. I do production work throughout the year that often requires an overnight render because it's effects-heavy or extremely high res for floor-to-ceiling displays or something of that nature. If I start one of these renders and it takes longer than six hours, the computer will flat out fail at the task. That, to me, is absolutely crazy. Gaming is not the only lengthy, high load situation that the owner of a quad core+dGPU notebook is going to encounter.

It happens all the time. I do transcode work for the company on my 12-core Hackintosh that can take up to two full days of processing to complete. A weekend-long shoot that generates 4TB of 6K footage which needs to be converted down to 1080p proxies for smooth post work - if I'm on the road, this limitation means I can't do it.


The 2016 MBP is presumably a professional computer. It should be able to run at full capacity unlimited and unending.
 
  • Like
Reactions: eulslix
Completely agree and tbh was never the behaviour on older macs. You could run them without a battery they just wouldn't run at full speed.

My Older 15" MBP`s exhibited this behaviour, 2008 and 2011 still do, takes a lot mind. Never had one shut off, just observed the charging light flip to red.

Q-6
[doublepost=1481309471][/doublepost]
To be clear, though, I have not done the full 6 hour stress to see if it will shut down in OS X or simply under clock to hit the 87w max on AC. I'd hope it just downclocks.

In theory it should throttle as either over temp or insufficient power can result in the CPU/GPU reducing frequency.

Q-6
 
That`s normal, many of the preceding MBP`s had the same issue, equally it`s just BS in 2016 for Apple to not ensure the power supply to the notebook is adequate for maximum load. Clearly an indication of Apple`s intended usage for the 15" MacBook Pro...

Q-6

You realize it's called a MacBook Pro, right? You can't have the word Book in the name if you make the device and its power brick heavier and take up more volume.
 
You realize it's called a MacBook Pro, right? You can't have the word Book in the name if you make the device and its power brick heavier and take up more volume.

Reality it`s down to the user need, highly likely such scenarios are uncommon, equally a larger power supply is all that`s required. I have seen the same behaviour on my older 15" MBP`s. For the most part majority of users wont push their MPB`s anywhere near hard enough to trigger this behaviour.

Q-6
 
  • Like
Reactions: jerryk
to be fair.... if you are maxing the CPU + GPU for more than 6 hours at a time, you should probably purchase a Desktop that can do this.

Moving files to and from that desktop isn't gonna be particularly convenient or fast... Why not just get a laptop that can do it, since just about every other vendor makes such laptops?
 
  • Like
Reactions: eulslix
Been stress testing a 15" 2.7/455/500 (using 100% CPU and 100% dGPU + full screen brightness) and in almost 3 hours the battery has dropped about 5%. Interestingly enough, I think the screen at full brightness is what pushes it over the line. If you turn the screen brightness down to about 50%, the system goes from averaging around 86-88 watts total (the power supply is 87 watts) to about 78-80. With the screen off you'd even get a bit better. So, with the middle model, if you keep the screen dim and/or completely off you'll probably be fine when using CPU and GPU at 100% continuously.

I should be receiving a 2.9/560/500 today and will be trying the same tests with it to see if the higher spec'd CPU and GPU actually have an effect (The CPU at least is the same TDP, so really shouldn't? Not sure about the GPU).

Thanks,
mbezzo
 
You realize it's called a MacBook Pro, right? You can't have the word Book in the name if you make the device and its power brick heavier and take up more volume.


But why not allow it to draw more than 87 watts through the brick then if it's not enough for some situations? All it takes is for any company to make a larger, "pro" 100w USB-C brick to supply all that power through AC in sustained high load situations.
[doublepost=1481314359][/doublepost]
Been stress testing a 15" 2.7/455/500 (using 100% CPU and 100% dGPU + full screen brightness) and in almost 3 hours the battery has dropped about 5%. Interestingly enough, I think the screen at full brightness is what what pushes it over the line. If you turn the screen brightness down to about 50%, the system goes from averaging around 86-88 watts (the power supply is 87 watts) for the entire system to about 78-80. With the screen off you'd even get a bit better. So, with the middle model, if you keep the screen dim and/or completely off you'll probably be fine when using CPU and GPU at 100% continuously.

I should be receiving a 2.9/560/500 today and will be trying the same tests with it to see if the higher spec'd CPU and GPU actually have an effect (The CPU at least is the same TDP, so really shouldn't? Not sure about the GPU).

Thanks,
mbezzo


Hmm, this is interesting and probably makes sense - the screen may be the tipping point. Back in the iPhone 6 Plus days if you were, say, playing a game or running GPS+music in your car plugged into a 1A charger, you could actually pinpoint the brightness level where it stopped losing charge.
 
Even more interesting! So I have had Intel Power Gadget open for this whole process, and it's at about 92% battery now at about 3.5 hours. So, it's been long enough I think to get a good idea of what's happening. From watching Intel Power Gadget, the CPU package has been averaging right around 35 watts and maintaining about 3.0Ghz. As soon as I quit the GPU stress app, the CPU immediately went up to about 40 watts (and as I type this has bumped up to 42 watts I presume as things cool down a little) running at 3.2Ghz. SO. Apple is clearly doing some behind the scenes "magic" to keep the total system power at roughly the total wattage of the power adapter. I'm very excited to see the CPU bumping up over the 35watt TDP "lower option" - which means they're allowing max performance for the CPU when it's not competing for power with the GPU. In reality, maxing out both CPU and GPU at the same time is gonna be pretty hard to do - possible of course, but a pretty fringe case.

Doubt I'll have time for the 2.9 test today, so that may have to wait for a bit.

thanks
 
Even more interesting! So I have had Intel Power Gadget open for this whole process, and it's at about 92% battery now at about 3.5 hours. So, it's been long enough I think to get a good idea of what's happening. From watching Intel Power Gadget, the CPU package has been averaging right around 35 watts and maintaining about 3.0Ghz. As soon as I quit the GPU stress app, the CPU immediately went up to about 40 watts (and as I type this has bumped up to 42 watts I presume as things cool down a little) running at 3.2Ghz. SO. Apple is clearly doing some behind the scenes "magic" to keep the total system power at roughly the total wattage of the power adapter. I'm very excited to see the CPU bumping up over the 35watt TDP "lower option" - which means they're allowing max performance for the CPU when it's not competing for power with the GPU. In reality, maxing out both CPU and GPU at the same time is gonna be pretty hard to do - possible of course, but a pretty fringe case.

Doubt I'll have time for the 2.9 test today, so that may have to wait for a bit.

thanks
I am willing to bet that when in bootcamp there is not the same safeguard
 
I am willing to bet that when in bootcamp there is not the same safeguard

No idea - I'll let someone else test that out. :) I bet this is probably an SMC level feature - I wouldn't be surprised to see this in Windows - but yeah, no real idea! :)
 
Reality it`s down to the user need, highly likely such scenarios are uncommon, equally a larger power supply is all that`s required. I have seen the same behaviour on my older 15" MBP`s. For the most part majority of users wont push their MPB`s anywhere near hard enough to trigger this behaviour.

Q-6

Why do you sign off posts with initials? You know we can see your username when you comment, right? Nobody is confused about who said what.
 
  • Like
Reactions: Brookzy
I am willing to bet that when in bootcamp there is not the same safeguard


In Windows, the computer unexpectedly shuts down with no warning. Happened to me twice and is what caused me to create this topic in the first place. It shut down on me at 7% and 12%, where it seemed like if it wasn't going to get that consistent boost off the battery, its only other option was to violently abort mission.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.