thermal throttling More details about the throttling issues of the 15" core i9 MacBook Pro, this time with Final Cut Pro X

[PeterBocan]

4 minutes ago, D13H4RD2L1V3 said:

Unless Apple is fine with having prematurely dying machines due to dead VRMs

That's for another class action lawsuit. Can you imagine working on a video project and the next day you realise that the Macbook Pro is dead because the VRMs just had enough? Abhorrent.

[D13H4RD]

2 minutes ago, PeterBocan said:

That's for another class action lawsuit. Can you imagine working on a video project and the next day you realise that the Macbook Pro is dead because the VRMs just had enough? Abhorrent.

Something needs to be done. 

 

From dusty Butterfly Keyboards to potential VRM failure. I really feel that this generation of MacBook Pro is slowly becoming the dud... 

 

Makes me happier that my old 2009 MacBook Pro still lives despite its status as an obsolete device 

[XenosTech]

I think stefan is confused about physics in general... How the hell can you say if something is drawing 150W that it puts out the same about in heat ? Whut ??

Pretty sure heat that comes from the transistors going on and off is due to the slight inefficiency of electrical circuits since a perfectly efficient circuit emits no heat at all. I mean what do I know, I'm just an assistant network administrator, what do I know about electrical circuits.

[PeterBocan]

34 minutes ago, XenosTech said:

I think stefan is confused about physics in general... How the hell can you say if something is drawing 150W that it puts out the same about in heat ? Whut ??

Pretty sure heat that comes from the transistors going on and off is due to the slight inefficiency of electrical circuits since a perfectly efficient circuit emits no heat at all. I mean what do I know, I'm just an assistant network administrator, what do I know about electrical circuits.

Well, I do remember the equation: Energy = Work + Heat. Where TDP just stands for the "Heat" part and not the work part. 

[XenosTech]

21 minutes ago, PeterBocan said:

Well, I do remember the equation: Energy = Work + Heat. Where TDP just stands for the "Heat" part and not the work part. 

Exactly and heat comes from work but that's based on how efficient it is at doing said work. Higher efficiency means less heat output regardless of how much energy was put into it at that start.

 

For a better analogy; If I drink 1 gallon of water and go play a game of basketball or go jogging, does my body sweat out a gallon of that water in the form of sweat or does my body sweat out a portion of that water ?

[Stefan Payne]

8 hours ago, mark_cameron said:

Apple are 100% to blame.

No, 70-80%, not 100

[Stefan Payne]

2 hours ago, XenosTech said:

I think stefan is confused about physics in general... How the hell can you say if something is drawing 150W that it puts out the same about in heat ? Whut ??

Never said that.

I said that if you put 150W in something, 150W will come out.

Since the CPU can not produce any movement energy or other forms, almost everything that goes in is converted to heat.

 

And I am not confused about Physics, I got indoctrinated with the Law of energy converson. Had to hear it every day multiple times.

Everything that goes in comes out. 

 

So I'm not the one confused about Physics, that part is something I really learned in school and hat to calculate efficiency and other stuff almost on a daily basis.

 

 

So why do you think that its not heat and other forms of energy?!

 

So your claim violates the laws of Energy Conservation!

Because Energy can only be converted in other forms of energy, there can not be any loss!

 

And what the normal people call "Power Loss" is not a loss of power it is a conversion to a form of Energy you do not want! And in most cases that happen to be heat (one way or another).

 

 

2 hours ago, XenosTech said:

Pretty sure heat that comes from the transistors going on and off is due to the slight inefficiency of electrical circuits since a perfectly efficient circuit emits no heat at all. I mean what do I know, I'm just an assistant network administrator, what do I know about electrical circuits.

Nothing it seems because its bullshit.

And as a Network administrator you don't have to know much about efficiency, conversion of energy and stuff like that, you are probably mostly a software guy and maybe need to know how long the cable can be before you need a repeater. But you have nothing to do with power consumption, conversion and other stuff. I did.

 

The heat doesn't come from transistors switching but from parasitic elements of semiconductor design like Capacitors and Inductors where you don't want them but they are there. 

Because every cable is a capacitor and an inductor!

And also CMOS Chips switch with an electrical short, wich is another reason why they consume power. But that amount could be negligable with modern designs and the other parasitic effects are more of a problem.

 

So could you guys please stop to claim that if 150W goes in there won't be 150W coming out?!

Because that is like claiming that the sky is pink!

And read up on that stuff before claiming that that doesn't happen...

[Commodus]

9 hours ago, mark_cameron said:

Exactly. Intel didn't force them to put it into the Apple machine. It's a processor for thicker laptops with liquid or other advanced cooling.

 

Apple are 100% to blame.

 

This is one of the worst computers ever put on the market.

 

Apple gets the award.

 

One of the worst computers ever on the market?  Er, no, that's just hyperbole.  It runs slower than it could, and there will likely be a software fix for this.  The build quality is excellent; the performance is still an improvement over its predecessors, and hard to find in other laptops this thin; the display and trackpad are outstanding; and now that the keyboard has been tweaked, it's at least solid for those who are fine with the key travel (I'd say the pre-2016 design is still nicer, but this is okay).  

 

I've used multiple machines that are way, way worse than this.

[TechyBen]

On 19/07/2018 at 12:02 PM, mr moose said:

Unless Intel's TDP rating is way off the mark (because  say apples cooling solution meets Intel's TDP rating) then it is apples fault.   That is why Intel have a TDP for their chips.

Nah, it's all Intels fault for not inventing magic chips, just like Apple who keep inventing magic products that "just work"...

 

/not sarcasm, but Apple Fan Mode.

[XenosTech]

3 minutes ago, Stefan Payne said:

-snip-

Heat is a form of energy...

Me being a network admin doesn't change the fact I did physics in school and that I'm a certified welder and electrician.

 

Know what I'm not even gonna bother, I deal with enough stupid people on a daily basis to add you to the list.

[Commodus]

Oh look, right on cue: Apple says software was to blame and has issued a fix.

 

There were some people who noted that it seemed more like it was a question of the temperature management rather than the nature of the design, and it appears that they were right.  We'll see how the fix works in practice, of course.  It's still likely to throttle at some point... just at a point where you'd expect any reasonably thin laptop to do it.

[Stefan Payne]

13 minutes ago, XenosTech said:

Heat is a form of energy...

Yes, I said as much and what are the other forms of Energy you were talking about?

 

Or were you wrong?

 

 

Quote

Me being a network admin doesn't change the fact I did physics in school and that I'm a certified welder and electrician.

Know what I'm not even gonna bother, I deal with enough stupid people on a daily basis to add you to the list.

Who you claim to be is not of interest as it can not be proven.

I could claim that I am a Physicist. What does that change in the validity of my or your postings?

 

If you would answer that with NOTHING, absolutely nothing!

You would be correct. And one Thing you should have learned about the Internet:
Nobody cares who you are or who you claim to be! Especially since it is not a provable thing. 

And whatever you had in School you seem to have forgotten. Wich isn't a problem. but claiming things that violate the laws of Physics is!

 

Just admit that you were wrong and everything is fine.

 

This is what I've learned:

https://de.wikipedia.org/wiki/Energieerhaltungssatz

(or for you in English: https://en.wikipedia.org/wiki/Conservation_of_energy )

 

Energy can only be converted to other forms of Energy and never be lost!

You however claimed that energy can be lost.

 

 

Technical English isn't my Strongpoint as I learned all that stuff in my Mother's tounge -> German.

And there are things I won't ever forget and this is one of the things!

[TechyBen]

On 19/07/2018 at 12:48 PM, leadeater said:

Don't worry according to the comment section on the article:

[comment on "update to OS will fix this in optimisation"]

A newer OS X version can fix this........  LOL!

Cargo cult style replies from fans right there. WOW.

[Stefan Payne]

1 minute ago, Commodus said:

Oh look, right on cue: Apple says software was to blame and has issued a fix.

 

There were some people who noted that it seemed more like it was a question of the temperature management rather than the nature of the design, and it appears that they were right.  We'll see how the fix works in practice, of course.  It's still likely to throttle at some point... just at a point where you'd expect any reasonably thin laptop to do it.

Yes, the Software could also have been the Problem or rather the Default Setting for the CPU and to let the CPU roam free.

 

You can fix that with Software as some users already have proven.

[PeterBocan]

5 minutes ago, Stefan Payne said:

Yes, the Software could also have been the Problem or rather the Default Setting for the CPU and to let the CPU roam free.

 

You can fix that with Software as some users already have proven.

You can never fix hardware issue with software update. You can only mitigate the stupid design. It's like trying to fix missing leg with, I don't know, VR headset. 

It's only a mitigation, not a fix.

[TechyBen]

58 minutes ago, Stefan Payne said:

Yes, the Software could also have been the Problem or rather the Default Setting for the CPU and to let the CPU roam free.

 

You can fix that with Software as some users already have proven.

If it was power management, and not fan curve, then no... other users did not prove/fix it.

If it was fan curve, then yes, Apple fixed it. I'm assuming it was power draw errors, pumping it up to too high a turbo for too long.

[reads link and Apples explanation]

Yep, I was right, they did not implement the power draw drop when thermal limits hit... IE, they told the chip "We got a beefing heatsink on, go wild!"

 

So not Intels fault. Not Adobe. Not users. Apples, and though a simple error, is amazing it was not tested for/picked up... but then again, sometimes VWs need to meet emissions tests and all that...  

[mark_cameron]

7 hours ago, Commodus said:

 

One of the worst computers ever on the market?  Er, no, that's just hyperbole.  It runs slower than it could, and there will likely be a software fix for this.  The build quality is excellent; the performance is still an improvement over its predecessors, and hard to find in other laptops this thin; the display and trackpad are outstanding; and now that the keyboard has been tweaked, it's at least solid for those who are fine with the key travel (I'd say the pre-2016 design is still nicer, but this is okay).  

 

I've used multiple machines that are way, way worse than this.

Name ONE computer that is more poorly designed than this US$3000 machine.

 

One.

[mark_cameron]

7 hours ago, Commodus said:

Oh look, right on cue: Apple says software was to blame and has issued a fix.

 

There were some people who noted that it seemed more like it was a question of the temperature management rather than the nature of the design, and it appears that they were right.  We'll see how the fix works in practice, of course.  It's still likely to throttle at some point... just at a point where you'd expect any reasonably thin laptop to do it.

That's nonsense.

 

This is a hardware issue. People have already made software changes and still found throttling.

 

You're paying for an i9 CPU

 

But you're getting a dud. No software patches are going to change a faulty hardware design and lack of cooling.

 

I also dismiss Apple's bench marks. They have used active cooling there. Like a water chiller. Or put the MacBook in a freezer.

[JoostinOnline]

4 minutes ago, mark_cameron said:

Name ONE computer that is more poorly designed than this US$3000 machine.

 

One.

Does the N64 controller count?

[79wjd]

54 minutes ago, JoostinOnline said:

Does the N64 controller count?

I don't see the problem

[TechGod]

17 hours ago, mark_cameron said:

Exactly. Intel didn't force them to put it into the Apple machine. It's a processor for thicker laptops with liquid or other advanced cooling.

 

Apple are 100% to blame.

 

This is one of the worst computers ever put on the market.

 

Apple gets the award.

Worst computers ever put to market? You're kidding right? 

[mark_cameron]

5 hours ago, TechGod said:

Worst computers ever put to market? You're kidding right? 

Nope.

 

This is a US$3000+ paperweight 

 

As far as I'm concerned. All it is good for.

[TechGod]

2 minutes ago, mark_cameron said:

Nope.

 

This is a US$3000+ paperweight 

 

As far as I'm concerned. All it is good for.

You're funny. 

[NinerL]

Seems to be fixed now.  https://www.macrumors.com/2018/07/24/throttling-fix-2018-macbook-pro-improvements/

[ImpulseRez]

On 23/07/2018 at 7:21 AM, Stefan Payne said:

you don't have any proof of that as the cross plattform benchmarks are pretty much impossible to do.

And the ARM Chips are developed for a total different usage pattern anyway, so it doesn't make too much sense to do that anyway...

Yeah that's my point. Anyone buying a top of the range Macbook pro is going to want to do some 'pro' things (e.g photo/video editing etc), which an ARM chip is totally unsuited to. No point having 'theoretical' performance from an ARM chip if it can't execute x86 code quickly (which all the main pieces of software are using!)