MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
1–10 of 39 posts
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#2Originally posted at MacRumors by winterny https://forums.macrumors.com/threads/optimal-cpu-tuning-sett...
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#3Originally posted at MacRumors by winterny https://forums.macrumors.com/threads/optimal-cpu-tuning-sett...
Thanks! Url changed from https://old.reddit.com/r/macbookpro/comments/91256u/optimal_..., which straight-up copies this.
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#4Seems crazy to spend so much on a machine, just have to do something that apparently can void your warranty if something goes amiss, to take advantage of the extra power you paid for.
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#5That is much more embarrassing than simple thermal throttling of the CPU. The thermal overload on PMIC chips are designed to handle malfunctions not regular usage.
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#6Phantom throttling has been a big issue in the desktop i9s too. I've ran into it overclock mine many times.
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#7On TWiT[0]. They had the keyboard up to 115f sometimes, some other videos on youtube have gone much higher.
I've also seen a couple of youtubers and the number is growing, returning their macbooks [1], [2], due to the performance and the number is growing daily.
I really think this issue is down to a few things.
1) The need from Apple to be thinner each iteration, is really costing them the ability to have better thermals and fans.
2) Putting an 14nm chip where a 10nm should be. Unfortunately Intel is having issues with that.
3) No vents at the sides or the back, only where the screen is. Poor design considerations when it comes to heat.
I'm going to be keeping onto my MBPr 2012 for another year. I want to see some of these issues being resolved.
I would not be surprised if this spurs on Apple to phase in Arm processors in 2020.
[0]: https://www.youtube.com/watch?v=6DcIP6FtuRk
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#8Originally posted at MacRumors by winterny https://forums.macrumors.com/threads/optimal-cpu-tuning-sett...
Thanks! Url changed from https://old.reddit.com/r/macbookpro/comments/91256u/optimal_... , which straight-up copies this.
I think it's the other way around - MacRumors post was July 22nd at 4:57pm while the Reddit post was July 22nd at 4:36pm GMT-7. (I can't tell if I'm being messed up by time zones though as that doesn't appear on mouseover for MacRumors)
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#9All powerfull cpus and gpus these days have what is called droop detection. When they detect undervoltage, they "stretch" the master clock.
Effectively, they skip clock cycles till voltage recovers in power supply capacitors.
Re: MacBook Pro throttling not a CPU issue, but rather power throttling of the VRM
#10Originally posted at MacRumors by winterny https://forums.macrumors.com/threads/optimal-cpu-tuning-sett...
An update from the Reddit comment that's not on the first page of the macrumors thread:
[Edited to add note: From messages others have sent me, it seems that this MSR is actually not Intel's default, but rather this is what Apple has been setting the MSR value to on all intel chips in Macbook Pro and Macbook Air for a number of years ... This is just the first chip to actually be capable of drawing enough power for this to matter. Previous chips just couldn't draw that much power, so they never hit the VRM limit before]