Intel is to Blame... It's True.

[throAU]

I think both you and Queen are a bit greedy here The problem of course is that the concept to turbo boost spoils people — they forget that its just an opportunistic feature and the nobody guarantees yo that the turbo will be maintained for prolonged periods of time. Of course, you can design a cooling solution that will maintain higher turbo rates indefinitely, but then you are essentially designing for a desktop CPU in a laptop. Might as well taken a 60W desktop Coffee Lake instead.

Greedy?

For wanting the same performance as others are getting out of the i9 for less money?
Or wanting an actual increase in performance over last year's model?

Perhaps the people not disappointed with this in the extreme are just giving an Apple a pass because fanboyism...

Look, i'm a massive Apple fanboy usually, but recent Macbook Pro releases are a joke.

 

[Queen6]

I think both you and Queen are a bit greedy here The problem of course is that the concept to turbo boost spoils people — they forget that its just an opportunistic feature and the nobody guarantees yo that the turbo will be maintained for prolonged periods of time. Of course, you can design a cooling solution that will maintain higher turbo rates indefinitely, but then you are essentially designing for a desktop CPU in a laptop. Might as well taken a 60W desktop Coffee Lake instead.

Anyway, ekwipt shows quite convincingly that the problem is not CPU heat dissipation but rather, voltage regulators overheating. I suppose you already saw this thread: https://forums.macrumors.com/thread...es-engineering-f.2128936/page-2#post-26276414 It seems that the default settings on these CPUs assume that the laptop can deliver steady 100W of power (not dissipate it, mind, thats a different thing!) which the MBP can't do. So you get power throttling, which is consistent with the results seen so far.

Other OEM's can manage and then some. The VRM throttling is a bigger concern as it doesn't bode well for longevity due the excessive heating & cooling cycles.

It's all pretty academic as even with the power regulated correctly the cooling remains inadequate...

Q-6

 

[throAU]

It's all pretty academic as even with the power regulated correctly the cooling remains inadequate...

What's the power brick size on the current 15" macbook pro? 100w? isn't USB type C charging max of 100 watts anyway? I remember thinking back when the Macbook Pro ditched magsafe and went to USB-C only that this was going to be a bit of a limiting move. 100 watts is fine for low power machines, but anything workstation-y needs a bigger power adapter than that to run beefy hardware in it - and now they standardised on USB-C for power, they're limited to the USB-C spec.

Even the old discrete GPU 2011 machines were marginal / over spec for the power brick. 97 watts of maximum consumption in the box with a 95 watt power adapter. It couldn't keep up but was close enough that battery could be kept full enough to last long enough under extreme load.

Will be funny to see how the i9 with discrete GPU goes when running both GPU and CPU heavily.

The charger simply won't be able to keep up, even if the VRMs inside the machine were OK - AND the cooling was adequate for the VRMs, CPU and GPU.

It's clear that Apple aren't interested in putting high end GPUs inside the Macbook pro - there simply isn't the power budget for it, irrespective of cooling. Now, even the CPUs are starved as well.

 

[haruhiko]

What's the power brick size on the current 15" macbook pro? 100w? isn't USB type C charging max of 100 watts anyway? I remember thinking back when the Macbook Pro ditched magsafe and went to USB-C only that this was going to be a bit of a limiting move. 100 watts is fine for low power machines, but anything workstation-y needs a bigger power adapter than that to run beefy hardware in it - and now they standardised on USB-C for power, they're limited to the USB-C spec.

Even the old discrete GPU 2011 machines were marginal / over spec for the power brick. 97 watts of maximum consumption in the box with a 95 watt power adapter. It couldn't keep up but was close enough that battery could be kept full enough to last long enough under extreme load.

Will be funny to see how the i9 with discrete GPU goes when running both GPU and CPU heavily.

The charger simply won't be able to keep up, even if the VRMs inside the machine were OK - AND the cooling was adequate for the VRMs, CPU and GPU.

It's clear that Apple aren't interested in putting high end GPUs inside the Macbook pro - there simply isn't the power budget for it, irrespective of cooling. Now, even the CPUs are starved as well.

Using more gas / power doesn't mean that it's more powerful.

 

[throAU]

Using more gas / power doesn't mean that it's more powerful.

??

More powerful CPUs and GPUs at higher clocks need more power.

This is due to this thing called the laws of physics.

Apple/intel have not broken those yet, much as Apple's marketing department might like to try and convince you otherwise.

 

[Queen6]

What's the power brick size on the current 15" macbook pro? 100w? isn't USB type C charging max of 100 watts anyway? I remember thinking back when the Macbook Pro ditched magsafe and went to USB-C only that this was going to be a bit of a limiting move. 100 watts is fine for low power machines, but anything workstation-y needs a bigger power adapter than that to run beefy hardware in it - and now they standardised on USB-C for power, they're limited to the USB-C spec.

Even the old discrete GPU 2011 machines were marginal / over spec for the power brick. 97 watts of maximum consumption in the box with a 95 watt power adapter. It couldn't keep up but was close enough that battery could be kept full enough to last long enough under extreme load.

Will be funny to see how the i9 with discrete GPU goes when running both GPU and CPU heavily.

The charger simply won't be able to keep up, even if the VRMs inside the machine were OK - AND the cooling was adequate for the VRMs, CPU and GPU.

It's clear that Apple aren't interested in putting high end GPUs inside the Macbook pro - there simply isn't the power budget for it, irrespective of cooling. Now, even the CPUs are starved as well.

Simply seems to be the direction Apple wants to follow...

USB C charging is a clear limit for more performance orientated notebooks. The current VRM throttling is little more than a joke at this point. It's also indicative of the loads Apple thinks it's users will apply to the system.

Power users should not have to go through a range of hacks to get the best if the system period. A $4000 notebook should be optimised by the OEM, the whole mess comes across as something Apple just threw together with little attention to details.

I am certainly rethinking, as the VRM issue is not correctable, yes you can mitigate, equally no matter what is done performance is significantly reduced.

When the GPU is added to the equation the situation only gets worse. This release of the MBP for the first time in years looked to be promising, only to be yet another MBP problematic iteration, with again the potential for reliability issues down the road.

Worst of all Apple will highly likely remain silent coming across as being little more than arrogant with little regard for the customer...

Q-6

 

[leman]

USB C charging is a clear limit for more performance orientated notebooks.

I've said it before and I'll say it again. There is nothing preventing anyone from designing a custom USB power delivery protocol that will deliver more then 100W. I doub't that Apple will ever do it of course, their laptops have always been limited to 85Watt. In fact, the 2016+ MBP is the first Apple laptop that can handle more power.

The current VRM throttling is little more than a joke at this point. It's also indicative of the loads Apple thinks it's users will apply to the system.

Oh don't be so dramatic. The VRM throttling is just because of misconfiguration and sloppiness. Its a bug that can be fixed. Once the config is fixed the MBP performs just as any other Coffee Lake machine of comparable size save for the three times larger gaming behemots.

Power users should not have to go through a range of hacks to get the best if the system period.

Power users that are interested in absolutely max performance should not get a MacBook Pro. This was the case 10 years ago and it is the case now. Plenty of machines out there that can deliver performance you need. I mean, you even got a larger machine because it suited your needs better than a Mac. I don't understand this urge to make the laptop something it has never been...

 

[Queen6]

I've said it before and I'll say it again. There is nothing preventing anyone from designing a custom USB power delivery protocol that will deliver more then 100W. I doub't that Apple will ever do it of course, their laptops have always been limited to 85Watt. In fact, the 2016+ MBP is the first Apple laptop that can handle more power.



Oh don't be so dramatic. The VRM throttling is just because of misconfiguration and sloppiness. Its a bug that can be fixed. Once the config is fixed the MBP performs just as any other Coffee Lake machine of comparable size save for the three times larger gaming behemots.



Power users that are interested in absolutely max performance should not get a MacBook Pro. This was the case 10 years ago and it is the case now. Plenty of machines out there that can deliver performance you need. I mean, you even got a larger machine because it suited your needs better than a Mac. I don't understand this urge to make the laptop something it has never been...

If Apple produced three legged donkey you would spend the rest of eternity to defend it. With your narrative switching with Apple's own errors. Wasn't so long ago that you were praising the maximum sustained performance of the MBP over others. Now Apple has dropped the ball again it's a different story...

Current problem is clear to all who are rational. People need to clearly understand such issues, not have them trivialised or brushed away.

Apple may or may not issue an update, although all the current bad press, piled on top of more bad press will likely be a strong driver. Even then to get the CPU & GPU under the power budget will have a significant impact on performance.

Now we see why no big announcement as Apple clearly knew these new MBP's presented performance issues, don't you think they even test them as the are just awesome by default...

Q-6

 

[maflynn]

Current problem is clear to all who are rational. People need to clearly understand such issues, not have them trivialised or brushed away.

There's no other excuse, other then Intel produced a very hot CPU, but as per their past behavior, provided Apple and others with advanced copies of the CPU so the manufacturers can design computers around those CPUs. At the very least Dell and Apple are two companies that chose not to make any internal changes and the results especially for the i9 is disastrous. I'm content with my purchase, and I have a work around, but that in no way removes apple's responsibility and screw up. I can see this causing a lawsuit

 

[AndyMacAndMic]

It is final. Apple admitted it is a bug. So it seems Intel is not to blame.

 

[The Mercurian]

It is final. Apple admitted it is a bug. So it seems Intel is not to blame.

Agree. Mods you can probably close this thread down

 

[DjanSeriyAnaplian]

It is final. Apple admitted it is a bug. So it seems Intel is not to blame.

OUR WORK HERE IS DONE.

 

[MBX]

Actually it is mostly Intel to blame.

Some years ago Intel communicated that 10nm would be ready for 2016. Apple hence designed the chassis with that in mind. We would not have these issues now. Instead Intel delayed their 10nm over and over.

But will Apple publicly bash Intel? No. So respect to Apple.

However Apple's decision to move away from Intel after 2020 with their own CPU's says a lot too. I'm sure Intel's delays and now to late 2019 for 10nm to be ready played a big role.

 

[JPack]

Actually it is mostly Intel to blame.

Some years ago Intel communicated that 10nm would be ready for 2016. Apple hence designed the chassis with that in mind. We would not have these issues now. Instead Intel delayed their 10nm over and over.

But will Apple publicly bash Intel? No. So respect to Apple.

However Apple's decision to move away from Intel after 2020 with their own CPU's says a lot too. I'm sure Intel's delays and now to late 2019 for 10nm to be ready played a big role.

Intel never communicated what power consumption 10nm would require.

The mental gymnastics of fanboys is reaching peak levels in this thread.

 

[TimmeyCook]

Intel never communicated what power consumption 10nm would require.

The mental gymnastics of fanboys is reaching peak levels in this thread.

When you lose an argument, call him a fanboy, that'll tell him!

 

[StellarVixen]

IBM got kicked by Apple for not being able to deliver a CPU with good performance per watt.

Similar fate is awaiting Intel. Just wait and see. Different reasons, but the similar outcome.

Apple wants powerful, but very power efficient CPUs. Intel cannot deliver that.

There will be someone who can.

 

[buran-energia]

There will be someone who can.

The new CPU most likely won't have Intel's instruction set, meaning most software will have to be recompiled or else it will run at a significant performance hit. Bootcamp won't work (if it will, Windows software will have the same issues: either it won't work or will run at a much slower speed). Virtual Machines like Parallels and VMWare won't work, etc. So, there are significant downsides when moving away from Intel or AMD.

 

[T'hain Esh Kelch]

IBM got kicked by Apple for not being able to deliver a CPU with good performance per watt.

Similar fate is awaiting Intel. Just wait and see. Different reasons, but the similar outcome.

Apple wants powerful, but very power efficient CPUs. Intel cannot deliver that.

There will be someone who can.

The problem is that Apple needs processors that scale a wide variation of performance and wattage. Except for AMD, I don't see who else could supply them, unless they want to mix suppliers, which I find unlikely. Even going with ARM for the low end machines, and X86-AMDs for the high-end, is not something I could see them do.

So it really boils down to one question: Is Apple ARM development going so well, that they can compete with Intel/AMD i9/Zen2? I find that even more unlikely.

 

[maflynn]

IBM got kicked by Apple for not being able to deliver a CPU with good performance per watt.

It was more then that. Yes Steve Jobs promised a 3GHz processor by the end of the year (or something to that effect), but IBM was unable to provide chips in quantities at the time, there was always a shortage of Macs because IBM was not able to produce enough CPUs. They also stopped focusing on improving the PPC, as their business plan changed. They were more focused in producing embedded systems and not upping the GHz or increasing performance. Apple was painted into a corner.

 

[The Mercurian]

It was more then that. Yes Steve Jobs promised a 3GHz processor by the end of the year (or something to that effect), but IBM was unable to provide chips in quantities at the time, there was always a shortage of Macs because IBM was not able to produce enough CPUs. They also stopped focusing on improving the PPC, as their business plan changed. They were more focused in producing embedded systems and not upping the GHz or increasing performance. Apple was painted into a corner.

IBM also got out of manufacturing entirely around that time did they not?

 

[maflynn]

IBM also got out of manufacturing entirely around that time did they not?

Both Moto and IBM were getting out, but I think IBM was still had fabricating plants during Apple's PPC time. It wasn't long after that time, they did.