MacInTouch Amazon link...
Channels
Apple, Products, Questions
That's what iFixit said, but looking at the photos, I see what appears to be that port on the other side of the motherboard - near the lower-right edge of the right-side fan assembly. ...
Can you be more specific about the exact location? I examined the enlarged pictures of the motherboard. The single socket on the underside shows a connector in Step 8. Since it's the only connector aimed towards the display I assume that's what it's for. Unless I missed something Step 8 also shows that every socket on the top side has a corresponding cable.
 


DFG

YouTuber Dave Lee looks at thermal throttling in the new MacBook Pro 15" with the highest processor option (i9). And it's bad. Waay bad. Forget about "Pro" workloads. If confirmed, it is disappointing that Apple just put a more powerful processor in without upgrading the cooling.

To be sure, the Dell XPS 15 suffers from the same illness, though probably to a lesser extent.

I would like for Dave to be more quantitative in his measurements of thermal throttling, and to compare the i7 and i9 processors, as well as the same options in the XPS15.
 



Ric Ford

MacInTouch
Much more about Apple's T2 processor, introduced in the iMac Pro and now in the 2018 MacBook Pro:
Duo Labs said:
Apple iMac Pro and Secure Storage
With the introduction of the iMac Pro by Apple in late December 2017, a number of brand-new features premiered on the macOS platform. While Apple had already introduced a dedicated security coprocessor with the inclusion of the T1 processor in the late 2016 MacBook Pro with Touch Bar, it served only a few specific tasks like driving the Touch Bar display and touch screen, reading fingerprints from the Touch ID sensor and storing fingerprint data in the T1’s Secure Enclave Processor or SEP.

In keeping with its limited system architecture role, the T1 Application Processor or AP was based on a 32-bit ARMv7 processor similar to the S1 AP used in the Apple Watch. Its dedicated watchOS-based OS is named BridgeOS, while the hardware identifier is iBridge.

The announcement of the iMac Pro made reference to the presence of another security processor but also mentioned a number of changes to system security. Touch ID would not be present and, while rumored to be an option, Face ID was also not present. Instead of focusing on mobile-friendly security features like touch or facial recognition, Apple instead used the next-generation T2 coprocessor to implement an overhauled boot security system, which it named Secure Boot.

Due to the increased reliance on a security coprocessor, Apple decided to give the T2 coprocessor a significant performance bump by basing it on the same AP core as the one found in the Apple A10 processor, a 64-bit ARMv8 processor also found in the iPhone 7 and 7 Plus. The T2 coprocessor is not exactly the same as the A10 processor since it incorporates only one of the T801x cores that make up the A10 package, but is a significant step up in capabilities from the earlier T1 AP. Apple gave the AP the model number T8012, references to which can be found throughout BridgeOS. A separate Secure Enclave Processor or SEP is also found in the T2, similar to that in the T1.

Given all of these changes, we wanted to explore how the T2 coprocessor was being used by Apple and how it currently fits into the larger system security model, as well as how this may evolve in the future. What follows is the first part of this exploration where we describe how the T2 coprocessor is used to implement Secure Boot on the iMac Pro, as well as comparing and contrasting this Secure Boot approach to those that have been present in Apple’s iDevices for a number of years.
 


(1) You should always have an active backup strategy; any drive can fail.
(2) By definition, any robust encryption should be resistant to recovery, right?
1: Yes, of course.

2: Agreed, but hardware encryption that ties the flash chips to a certain controller changes recovery from "annoying" to "nearly impossible". If you are using FileVault (or some other similar system) without encryption that keys the flash to the controller, then you can (theoretically) connect the flash to a new controller, provide your FileVault credentials and get access. As long as the chips aren't soldered to the controller's board, the process shouldn't be all that difficult. By tying it to a specific controller chip, that approach no longer works.

This difference is most important for something like an iMac Pro, where the flash chips are on separate modules that can be removed and replaced.

On something like a MacBook Pro, where the flash and the controller are soldered to the same board, it probably doesn't matter, because you'd need a lot of expensive equipment to remove the chips and attach them to a new controller. But even so, this is something a data recovery service could do if the chips weren't cryptographically keyed to the (presumably defective) controller chip.
 


looking at the photos, I see what appears to be that port on the other side of the motherboard - near the lower-right edge of the right-side fan assembly.
After asking this in a comment on the tear-down page, I was corrected. That connector is where one of the TouchBar cables connects. This wasn't clear from the photos. So there is no SSD-recovery port, it seems.
 



I can't help but think the T2 processor is primarily to stem the creation of hackintoshes. Throw some custom silicon in it, require it for basic functions like booting, and hackintoshes go bye-bye.
 



I can't help but think the T2 processor is primarily to stem the creation of hackintoshes. Throw some custom silicon in it, require it for basic functions like booting, and hackintoshes go bye-bye.
Or it could be about system security, which, frankly, I value quite highly. Having another processor with a different OS, designed with secure operation in mind, makes a lot of sense.

That's not saying you're not correct, but, frankly, just how much does the hackintosh market hurt Apple's sales?
 


Ric Ford

MacInTouch
The Geekbench guy did some developer workload (compile) tests to look at throttling of the 2018 MacBook Pro (but didn't have access to an i9 CPU to test):
John Poole/Primate Labs said:
MacBook Pro (Mid 2018) Throttling
...
Conclusions

So what’s going on here? Why does this test not replicate the throttling seen in other tests? Part of the issue is the tests themselves. Premiere uses both the CPU and the GPU, while Geekbench only uses the CPU. If the GPU contributes significant heat, then that will cause the CPU to throttle more aggressively. It’s possible the decrease in performance observed in Premiere is due to a combination of new AMD GPUs with new Intel processors, or to the new AMD GPUs themselves.

My recommendation? If your work doesn’t involve long-running tasks that are CPU- and GPU-intensive (such as Premiere) then the new MacBook Pro should provide a considerable increase in performance. Otherwise, it might be wise to wait until more performance data is available.
 


I presume that these new MacBook Pro models won't be able to run Sierra. (I'm still leery about High Sierra and APFS.)
 


Ric Ford

MacInTouch
Here's a fun software toy for CPU testing:
Intel said:
Intel Power Gadget

Intel® Power Gadget is a software-based power usage monitoring tool enabled for Intel® Core™ processors (from 2nd Generation up to 7th Generation Intel® Core™ processors). Intel® Atom™ processors are not supported. It is supported on Windows* and Mac OS X* and includes an application, driver, and libraries to monitor and estimate real-time processor package power information in watts using the energy counters in the processor.

... Traditional methods to estimate power/energy usage of the processor has always been a cumbersome task that included special purpose tools or instrumentation on the platform along with third party equipment. The motivation for the tool was to assist end-users, ISV’s, OEM’s, developers, and others interested in a more precise estimation of power from a software level without any H/W instrumentation.
 


Ric Ford

MacInTouch
Still more on the Touch Bar MacBook Pro keyboard problems and changes:
iFixit said:
Anatomy of a Butterfly (Keyboard)—Teardown Style

... Since Apple informed their service providers that the membrane is “to prevent debris,” we’re inclined to think any change in noise level really is just a secondary feature.

... Apple has a proven track record of failure for these keyboards. They’re being accused, by way of several class-action lawsuits, of knowingly selling failure-prone keyboards. Apple may claim that they design products to last—and that designing for repairability compromises the durability of a device—but this keyboard misadventure belies those points. If a single grain of sand can bring a computer to a grinding halt, that’s not built to last. If said computer can only be fixed by throwing half of it away and starting over, that’s not built to last. We’re definitely excited to see improved protection on these machines—consumers deserve it with the prices they’re paying.
 


Another analysis of the throttling issues with the 2018 MacBook Pro:
ItsQuadPod/Reddit said:
What's In a TDP? Examining the Thermal Issues in the New MacBook Pro

While the old CPUs could only produce up to 45W at full boost clocks, the Coffee Lake chips will now draw up to 100W at their own. This amount of power is far beyond what the MacBook Pro was designed for, and will only reasonably be held for a short amount of time before the temperature spikes way up, at which point the CPU should throttle back down to its base clock speed where it produces 45W and the cooling system can keep the clock speed steady. However, the fans in Macs take notoriously long to ramp up so this does not happen and clock speeds tend to rapidly spike up and down without much control right now. Additionally, in loads where the CPU and GPU must work at the same time clock speeds fall even more than before thanks to those two extra cores and performance may drop below the old levels.
This makes it sound like Apple might be able to do some tuning on the fan speeds to at least partially mitigate the problem.

My impression (based on what information is out right now) is that the keyboard issues should (mostly? completely?) be mitigated with the membrane, and if your use case is such that you'll be working both the CPU and GPU hard on a regular basis, you're better off with an iMac or Mac Pro. Perhaps an external GPU might help mitigate this on the MacBook Pro, but you lose laptop portability at that point.
 


I can't help but think the T2 processor is primarily to stem the creation of hackintoshes. Throw some custom silicon in it, require it for basic functions like booting, and hackintoshes go bye-bye.
I might get that paranoid if the rumored (yet again) upgraded Mini has a T2 chip, but I'm guessing it won't. Aside from a couple of egregious cases of people hawking systems with Apple components, Apple has pretty much left the clone/Hackinstosh market alone since August, 1997, when they bought out Power Computing. I think the management at Apple is sensible enough to realize that those people willing to literally hack around to build a Hackintosh don't represent much in terms of lost sales to Apple — most would just build Windows or Linux machines if their access to the OS were cut off.

A more serious discussion topic might be whether the average user needs or wants the level of security that data-at-rest encryption provides, at the cost of time-consuming and even potentially disastrous experiences if hardware damage occurs. I suppose Apple knows the internal component IDs used for hashes in T2-equipped models and could custom-burn a new T2 for owners of record who'd had theirs destroyed somehow... but somehow I doubt that's ever going to happen.
 


I can't help but think the T2 processor is primarily to stem the creation of hackintoshes. Throw some custom silicon in it, require it for basic functions like booting, and hackintoshes go bye-bye.
I doubt it. In order to make that work, Apple would have to modify macOS to refuse to boot on any Mac without a T2 chip. Maybe that will be possible in a few years, but right now most Macs (all but the newest MacBook Pro and iMac Pro) don't have the chip. Apple isn't going to make an OS change that breaks their entire user base. And even if they wanted to, could you imagine the lawsuits that would ensue if the next macOS couldn't boot on any Mac other than the two newest models, including ones still in production?

Keep in mind also that people were posting similar fears when the first Intel Macs were announced, because those first prototype Intel Mac developer workstations all had TPM chips on the motherboard. Everybody was afraid that Apple was going to lock the OS to TPM-hosted credentials. It didn't happen. Quite the opposite, the Intel Macs that actually shipped didn't even have TPM chips.

In other words, while it is theoretically possible that Apple might do what you're afraid of, I wouldn't lose any sleep over it. They've had this ability for a very long time and have not decided to do anything so far. The presence of the T2 chip without any additional facts doesn't sound like a strong enough argument to conclude that they're changing their opinion after all these years.

Of course, if I'm wrong and Apple does decide to lock down macOS like this, I think you can rest assured that the Hackintosh community will find the relevant parts of the OS and disable and/or replace them. The iOS Jailbreak community has proven that there is no such thing as a lock that can't be picked.
 


I can't help but think the T2 processor is primarily to stem the creation of hackintoshes. Throw some custom silicon in it, require it for basic functions like booting, and hackintoshes go bye-bye.
I genuinely can't imagine that Apple cares that much. The number of sales lost to Hackintoshes has got to be a rounding error to Apple; the absolute number can't be that large, and some significant fraction of those who currently build them probably wouldn't be buying a new Apple machine if they couldn't. Given the engineering cost in designing the T2 and its subsystem, building support for it into the OS, and eventually including a chip in every Mac the company ships, I'd be shocked if that could possibly make financial sense if its goal was the elimination of Hackintoshes.

On the other hand, it's doing some very cool stuff, and may enable even more in the future by virtue of being a fairly powerful CPU with its own micro-OS. I don't think anyone knows clock speed on it, but with a single A10 comparable core, it could well be as fast as one of the cores in the 2016 MacBook (or one core of a 2011 or 2012-era i5 laptop CPU), which is really impressive if you think about it.

As a privacy advocate I'm quite pleased with the cryptography that Apple has gone all-in on. I likewise doubt that Apple's aggressive security is enough of a sales point to make the engineering on it break even from a financial perspective, since consumers mostly just don't get or care about security, so I appreciate its inclusion all the more.
 


Ric Ford

MacInTouch
I had forgotten about this hidden limitation in the 2016-2017 Touch Bar MacBook Pros that is resolved in the 2018 redesign:
The Verge said:
Apple’s new 13-inch Touch Bar MacBook Pros now have four full-speed Thunderbolt 3 ports
It’s a small but useful update to the earlier 2016 and 2017 13-inch models, which only offered full-speed data on the two left-side ports, while the two right-side ports were limited to slower PCI Express bandwidth. The reason for the shift is simple: the original Intel chips on the old 13-inch MacBooks only offered 12 PCI Express lanes, which wasn’t enough for the full 40Gbps transfer speeds on all four ports.

The new eighth-generation chips on the 13-inch model now offer up to 16 PCI Express lanes, which is enough for all four ports to operate at full speed, just like its larger, 15-inch sibling.
 


I can't help but think the T2 processor is primarily to stem the creation of hackintoshes. Throw some custom silicon in it, require it for basic functions like booting, and hackintoshes go bye-bye.
It's possible, obviously, but will take a long time for this hardware to filter through the entire product line. So it seems several years off - at minimum - before this could become a requirement to run macOS.

More realistically, hackintoshes simply won't run features that are important for security (a limitation that currently applies to FileVault on boot drives).

And I doubt that Apple is particularly concerned about hackintoshes as a hit to their market - it's marginal.

I think the T2 should be seen as a trade-off for customers (business, government) that have stringent security requirements - for whom the added security of the T2 chip integrated with the OS may be a significant benefit. Or to formulate as a question, does anyone know of any existing or pending security guidelines or standards that may require the type of lock-down a T2-type chip can enable? I wouldn't be surprised if such requirements drive a wave of upgrades in future, to Apple's benefit.
 


Another analysis of the throttling issues with the 2018 MacBook Pro:
This makes it sound like Apple might be able to do some tuning on the fan speeds to at least partially mitigate the problem....
Given the information out there now, that sounds like a possibility.

One thing I've been using on my 2012 15" is TG Pro (from Tunabelly), since it allows you to control when and how quickly the fans come on. It helps to keep the Mac cooler than it would be on its own. I do hear the fans more often, but since the 2011 17" died, I've been using it to (hopefully) avoid thermal damage.

I've been considering replacing the 2012 15" with the 2018 myself, but the thermal issues gave me pause. The new information further up in this thread is encouraging.

(No affiliation with TG Pro, by the way, just someone who has used it for 4 years now.)
 


I am now wondering about the thermal issues with the i9 chip, since we are now seeing more reports and not depending on a single data set. I do not think tweaking the fans is going to help that much, since some testing was done with the fans at full power.

I am looking forward to a battery of tests being run on all 13" and 15" systems. My suspicion is that the issue is limited to the 15" i9, but we need solid, consistent tests. We should also see what triggers the problem. Is it only when all 6 cores are at full speed? I think the Final Cut Pro and Premiere tests are the most useful, since they represent actual use of many people who are interested in the i9 configuration, and they run the CPU at full speed.
 


I might get that paranoid if the rumored (yet again) upgraded Mini has a T2 chip, but I'm guessing it won't.
The T2 chip is, among a few other things, Apple's default SSD for Macs. (Well, technically, it is the SSD controller, but with Apple's embedded and flash NAND only board designs, pragmatically it is the SSD, as far as the system is concerned).

Apple's march to making all Macs primarily SSD-based should be relatively transparent at this point. The only hold-up is SSDs being "affordable enough" for the entry-level models. There are no Mac laptops with a hard disk drive at this point, and the top end of the desktop line-up is SSD only.

Apple wants to put SSDs in all Macs. T2 is an SSD. Therefore, don't be surprised when some later variant of the T2 appears in all Macs somewhere down the road. If flash memory cost the same $/GB as rotating platter storage, Apple would have already made the move across the board. The T2 will keep Apple's SSD prices a bit higher than mainstream.

I think it is doubtful that Apple will shift the T2 boot policies so that the singular internal Apple SSD is the only possible drive the Mac can boot off. As long as the Macs can boot off of non-Apple SSDs, then the path to making a non-Apple system work is relatively viable.

Thunderbolt pragmatically allows a path for low-level boot drives to be added to a Mac system, so the concept of additional drives other than the singular Apple boot drive isn't going away any time soon. I think Apple is going to optimize along the default, "out of the box" experience and make it a secure boot with just Apple stuff, but they aren't going to kill off the additional value-add of expanding the system in all usages. (That isn't internal components, that is any possible expansion.)

I think the management at Apple is sensible enough to realize that those people willing to literally hack around to build a Hackintosh don't represent much in terms of lost sales to Apple — most would just build Windows or Linux machines if their access to the OS were cut off.
I'd guess the percentage is down in the 1% range. That stays manageable as long as Mac sales are growing (even at a slow rate close to flat). If Mac sales fall backwards significantly, that would be a problem.

It is unlikely most of those folks won't largely 'shift' to Linux/Windows. A bigger 'war' of measure-vs-counter-measure would likely escalate.
A more serious discussion topic might be whether the average user needs or wants the level of security that data-at-rest encryption provides, at the cost of time-consuming and even potentially disastrous experiences if hardware damage occurs.
Hardware damage isn't the most likely issue. The majority of users have mobile computers. Those get lost, stolen, compromised, etc. at a far higher rate than some drive hardware failure.

The "time-consuming" cost if you have a back-up is a restore (which is a couple of hours and close to free in cost). That isn't that high.

As SSDs get in the zone of triple- and quad-level charge storage in a single cell, the low-evel recovery process is going to get harder as the schemes to encode and store the data get more complicated. If the SSD controller gets borked and the metadata gets fried, there isn't going to be a whole lot of recovery, even if that data is not encrypted.

Modern SSDs don't do 1-to-1 block mapping from logical address to physical. Even without encryption, all the data is encoded in some way. The notion of "I'll get a magnet and read it straight off the media" to recover the data overlooks what is actually under the covers in modern SSDs.

I suppose Apple knows the internal component IDs used for hashes in T2-equipped models and could custom-burn a new T2 for owners of record who'd had theirs destroyed somehow... but somehow I doubt that's ever going to happen.
I doubt they have a record. If they did, it would be subject to a 'subpoena' in whatever country they operated in. Also eliminates the chance a rogue employee will steal the keys and significantly damage their security reputation. If not there, it can't be stolen.
 


Apple’s new 13-inch Touch Bar MacBook Pros now have four full-speed Thunderbolt 3 ports
The reason for the shift is simple: the original Intel chips on the old 13-inch MacBooks only offered 12 PCI Express lanes, which wasn’t enough for the full 40Gbps transfer speeds on all four ports.
First, Thunderbolt is 40Gbp/s from the Thunderbolt controllers. That is the Thunderbolt speed; not the PCI-e speed. Second, x4 PCI-e v3 is 32Gb/s. Thunderbolt does 40, so that can deliver 32Gb/s in a timely, relatively low-latency fashion. If delivering other protocols, you need to be faster than what you are hauling/delivering.

A Thunderbolt controller only takes an x4 PCI-e v3 connection. Each controller does two ports, so you only need two controllers for a 4-port system. Two Thunderbolt controllers just need x8 lanes. 12 > 8, so this explanation is more than a bit odd. There is no discrete GPU in the 13" model, so not sure what would be consuming the PCI links off the CPU. The rest of the I/O would be hanging of the PCH chip. I suspect it is not the number, but the default grouping the CPU allows (but that's odd, too, because 3 by 4x would be a natural way to group 12. The only problem would be if it was x4 and then couldn't chunk another x4 out of the remaining 8).

I'd suspect a board layout problem and/or costs as being the problem far more than lack of gross oversupply of PCI-e lanes. For example, if Apple used an x2 connection to the Thunderbolt controller, there would be lower PCI-e bandwidth to that controller. That doesn't mean the controller couldn't approach 40GB/s if fed with two DisplayPort feeds plus that x2 feed.
which is enough for all four ports to operate at full speed, just like its larger, 15-inch sibling.
Actually, it is the 15" model with a discrete GPU than has the larger bandwidth restriction. The GPU is soaking up x8 (probably not all x16) of the PCI-e lanes coming out of the CPU package. The lack of a dGPU in the 13" model means there are more unused bandwidth, not less.
 


I am now wondering about the thermal issues with the i9 chip...
The root issue isn't the i9 chip package, it is the overall system. The natural Apple Store selection path to an i9 MacBook Pro 15" bundles it with the highest GPU also. (There is another BTO path by selecting the lowest-price system and bumping up the CPU but not the GPU.)

A 555X versus 560X probably would completely solve the problem, but should provide a small bit of "headroom" inside the system. If the heat pipe design stayed the same, the heat pipes for the CPU and GPU are coupled, to provide each with outflow access to each of the two fans, but on max load from both, that should be a coupled throttling issue on both.
We should also see what triggers the problem. Is it only when all 6 cores are at full speed? I think the Final Cut Pro and Premiere tests are the most useful, since they represent actual use of many people who are interested in the i9 configuration, and they run the CPU at full speed.
Concurrent max CPU and GPU is quite likely a trigger. The other two 15" CPU offerings - i7-8850H and i7-8750H - have TDP configurable down to 35W (Apple apparently is not running in that mode, but they can be stepped down). The i9 model can't. It is probably just barely squeezing under the 45W TDP limit. If you throw in a GPU upgrade, which also has some TDP scope creep, and don't change the physical cooling system parameters at all, then this result wouldn't be surprising at all.

Most of the rest of the configs would work OK, but max CPU + max GPU isn't the focus of the design. Apple dropped the old 15" design, but if Apple had a "desktop replacement" oriented model also in the line up, that would have been a more natural fit for the i9 plus as big (hot) of a GPU as they could fit.
 


I genuinely can't imagine that Apple cares that much. The number of sales lost to Hackintoshes has got to be a rounding error to Apple; the absolute number can't be that large, and some significant fraction of those who currently build them probably wouldn't be buying a new Apple machine if they couldn't.
Not just that, but that small number of Hacintosh users are socially well connected, often Apple customers with another device, or possibly future Apple customers. Alienating them by stomping their small, active community would likely snowball and hurt Apple way more than ignoring it would.
 



Not just that, but that small number of Hacintosh users are socially well connected, often Apple customers with another device, or possibly future Apple customers. Alienating them by stomping their small, active community would likely snowball and hurt Apple way more than ignoring it would.
I built a hackintosh on a self-built homebrew machine similar to a pro model and installed Mac OS X 10.6 through about 10.9. It was a lot of bother and a good time-waster as opposed to other time-wasters.

We own eight operable Macs, as it is. The most efficient way to be a hackintosh owner/builder is to own a Mac, so you can easily download the OS installer. So, I doubt Apple has really lost much business.
 


Amazon disclaimer:
As an Amazon Associate I earn from qualifying purchases.

Latest posts