Apple releases software fix for MacBook Pro slowdown
sixcolors.com
sixcolors.com
[0] https://www.dropbox.com/s/y30x13mi6ozhg0k/10.16.3S-Fix.png?d...
EDIT: If you'd like more information than just heresay, watch some of Louis Rossmann's videos on macbook heating problems (the video on tantalum capacitors is particularly interesting).
That seems damn high to me, but then I'm most familiar with desktops and servers, where you can cool far more aggressively. So is long term ~92°C hard on CPUs?
0) https://www.cnet.com/news/apple-macbook-pro-speed-throttling...
So maybe someone could check exhaust temperature on a 2018 Macbook Pro with i7 under heavy load. I get that the exhaust ports are on the back, near the display hinge. But that still might be iffy, for those who cross their legs.
Any idea what "digital key" could mean here?
I got my bumper in the mail, but never needed or used it.
Exactly. It wasn't that big of a deal.
I mean, we haven't even gotten into how a bunch of other phones which were on sale at the time exhibited the exact same behavior.
>As mentioned in our detailed review, the XPS 15 9560 with the Core i7-7700HQ processor is prone to two types of throttling:
Thermal throttling of the CPU or GPU (generally the CPU) when temperatures get too high
Voltage Regulator Module (VRM) throttling caused by it getting too hot and being unable to deliver enough power
You have to wonder why a well known laptop that can't run a mainstream i7 without severe throttling is a story that couldn't gain any traction.
>The service manual for the Dell XPS 15 9570 is now available and it reveals something that we wished would improve in this generation — that the XPS 15 9570 carries over the same 2-pipe heatsink assembly from the previous generation. Not just that, there is also no indication of improved thermal dissipation from the Voltage Regulator Modules (VRMs) implying that we could be in for some serious throttling issues.
The Dell XPS 15 while earning accolades for being a very capable multimedia laptop that can also game, has also earned the reputation for throttling heavily under load. Both the earlier XPS 15s suffered from both thermal and power-limit throttling. The only way to fix this was to repaste and undervolt the CPU and also apply thermal pads on the VRMs to drive away all the excess heat.
Unfortunately, it looks like Dell has chosen to retain the existing heatsink design for the XPS 15 9570. Compounding throttling fears is the availability of a Core i9-8950HK variant.
https://www.notebookcheck.net/The-Dell-XPS-15-9570-might-inh...
You would think that after two years with the same issue, the mainstream tech press would pick up on it if the issue is throttling and not just chasing clickbait stories.
I think the official explanation seems pretty likely. If the flaw had been present in the machines used for benchmarking than either their benchmarks didn’t catch the issue (unlikely) or the benchmarking numbers were fabricated which also seems unlikely as it would invite an infeasible amount of legal risk for an organization of apples size and mindshare ...
Clarification -- I meant to say -- 'It would be _very_ concerning if the benchmarks they ran for engineering/marketing production were insufficient to catch this issue on a system configuration in which the issue manifests.'
Which given the track record of quality issues they have had, is possible.
Exactly. I work in software engineering for cars and even for things that are not mission critical or life-threatening, the engineering and quality process is insane because so much money is on the line. I wouldn't be surprised if macbook pros are worth more in profit, or at least revenue, than a car line and should therefore see an equal engineering process.
I can't understand how the firmware was released without the proper key, unless there was a bad merge or they dropped a pilot firmware that was testing without thermal management... maybe a clue there is a desktop in the works and the desktop firmware made it onto the macbook pros? Either way, like you said, their quality process has a major deficiency.
Same story with the keyboard that was a fundamental enough design flaw to include 2 design bandaids in as many years.
Does that really matter though? I was just discussing this with someone over the weekend... his thesis was that people buy Apple because of branding, which might be true as far as the majority of customers are concerned... But for people like me (powerusers), MacBooks are simply the best possible choice. Everything else is either adware (Windows 10), or not usable (Linux laptops lack deiver support (last I heard) You can’t install software on Chromebooks, Windows 7 doesn’t have a Unix subsystem). Not to mention that Apple hardware is beyond superior, and the software mostly just works (things like wi-fi, backups, sharing, ...)
I am running Windows 10 Pro on a Surface. I haven’t seen a single ad, anywhere. If that’s your only reason let me set your mind at rest. Surface is a better Mac than a Mac at this point. That’s why I upgraded my old MBP to one rather than a new MBP.
But by all means ignore my advice and pay through the nose for a device that is blatantly not fit for purpose, I'll be over here enjoying the superior computing experience that Steve Jobs would have wanted me to have!
Yes, the promotional app stubs are advertisements, but they're easily removed and don't re-appear. In my opinion, removing the tiles you don't want is closer to an act of operating system "customization" than the user experience with ads on the web.
I had to change a registry key to stop it from installing Facebook and Candy Crush on its own.
But it's true that you better buy a laptop that you know has good support. eg: almost any dell, almost any asus, any thinkpad.
Currently, my XPS 15 with Xubuntu doesn't function after coming back from suspend and I haven't bothered to dig into why. I don't mind, but it's absolutely not a problem I experienced with windows on the same hardware or any mac I've ever owned.
I mean, I still love it and these aren't life or death issues, it's just something fundamental about the nature of Linux in an ecosystem of proprietary hardware specs and commercial drivers. It will always be a fact of life. Just like Linux package management, because the software is open source, will always be better than on Windows or a Mac. That's a fact of life that can't really change because of the nature of _their_ ecosystems.
No harm in admitting the differences, but good for you if everything works out of the box!
Dismissing windows 10 as merely "adware" is inflammatory to the point of meaninglessness. It's really the kind of rhetoric I would expect on /g/ not here. The OP has also been rightly called out by others for his broad dismissal of linux.
What's more important is fixing and moving on to the next problem.
Technologies are so complex nowadays that it is becoming virtually impossible to produce anything without issues.
Batterygate recently, or bendgate?
It's a slab of glass. What kind of person thinks that's a good idea?
Does that mean the whole unit will now run hotter before it starts to throttle? How is this going to impact some of those bad solder joints MacBooks are famous for?
It's not about what you do when the temperature if at its peaks and the fans are going at full speed, it's about how you avoid getting there.
>How is this going to impact some of those bad solder joints MacBooks are famous for?
When did MacBooks became famous for "bad solder joints"?
1) Louis Rossman has a vested interest in drumming up attention for his videos and has, on multiple occasions, used false statements and misleading practices to do so. Just like in the initial release video of the Touch Bar MBPs where he used an unshielded USB-A to USB-C cable and then complained that WiFi wasn't working. When used with a shielded cable, the issue went away and yet he never issued a correction or retraction. The issue was with the adapter he used, not the computer. Which leads directly to...
2) The issue stated in that video that required reballing was an issue with nVidia chips and affected all laptops that used that platform, both PC and Mac. To suggest that it was an engineering issue on Apple's part is disingenuous and only further reinforces the idea that he's not making these videos to be helpful but to drive traffic to his channel and repairs to his store.
3) Just look at how he used the Linus story to get the repair done at his shop. He has a pattern and, unlike what most people think, it's not altruistic. It's all to benefit him. I don't blame him but let's not pretend that he's doing this because he cares about users or the "right to repair" movement. He cares because he can mislabel and misconstrue the positions of the people against it to benefit his business. He'd love it if Apple was forced to provide him schematics and docs. He's in it for himself.
(There is a remarkable difference in mindset for the majority of MacBook and PC laptop owners. PC laptop owners will often just ditch their machine if it stops working. It's weird to say but I see it all the time.)
Now I don't doubt that most of what Louis says is correct. But I also appreciate much of it from Apple's perspective. If everyone in the independent repair industry was at least as competent as Louis is today, Louis would have a point. But they're not, so he doesn't. Besides, Apple's intransigence creates the market which Louis takes advantage of; he shouldn't complain.
Oh and I keep getting irritated every time Louis bangs on about those damn "refurbished" iPhone screen assemblies. He should know full well that the replacement glass used on these refurbished assemblies is cheap fragile non-Gorilla junk. I know multiple people who have had their iPhones independently repaired only to have the screen break again in a matter of days or weeks, under the most innocuous of circumstances. We are being scammed, and Louis defends the scammers.
Otherwise it's a great channel.
I know multiple people who have had their iPhones independently repaired only to have the screen break again in a matter of days or weeks, under the most innocuous of circumstances.
I realize this will be unpopular, but my observation about people who walk around with perennially broken glass on their phones, is that some people can't keep nice things. If your lifestyle is that rough, either use a completely enclosing polycarbonate case, or get a more robust 2nd phone. Expecting a computer made largely of glass to stand up to roughhousing contradicts common sense.
He pretty much does.
"So what they do is they take this screen ... and they replace the front glass layer. Most of the time the only thing wrong with the screen is the front glass layer is cracked. They put a new glass on it in China and then they sell it to us in the United States as a refurbished iPhone screen ... That is not a counterfeit."
https://www.youtube.com/watch?v=HkPUoUJPJAQ&t=1m55s
I understand why Louis gets worked up about this particular issue, but he's just flat out wrong. These refurbs are dodgy as fuck. The assemblies as sold are counterfeit.
> some people can't keep nice things
Absolutely. I don't disagree, but this doesn't describe the people I spoke about. Accidents happen to all of us, even when we're particularly careful. I would never expect a glass computer to survive rough treatment but I do expect that when it's repaired with a "refurbished Apple screen" that it is made of substantially the same material as the original.
Which is exactly why Apple's distinction was the logo on the product. If it's not the exact same product, it shouldn't be able to bear their logo. Judges deciding that it doesn't matter because users will never see the logo are missing the point, imo. It only opens the door to all kinds of 3rd party parts and repairs where it's not the exact same product but you're paying for it as if it is.
This is exactly the point that I feel like I constantly have to get across to people that support him. Louis is incredibly smart when it comes to electronics repair. It's exactly the reason that I know, for certain, that he's being disingenuous. He's too smart and too knowledgeable in most of these topics to gloss over these issues the way that he does unless he has an ulterior motive and it's clear that he does. Apple's fight against "Right to Repair" (which is misleadingly named anyways) isn't about people repairing their own devices, it never was. It's about preventing repair shops from doing those repairs and giving Apple a bad name because someone sees an iPhone where the screen fell off the front ("where the front's not supposed to fall off!").
I have had a Mac since the 512K days and they have always shipped with commodity components and Apple has never had some magic pixie dust that made them fail at lower rates than if they were in a PC. Nor have they ever individually tested every machine to guarantee it works.
Apple has and will always ship with broken video cards. And broken displays. And broken CPUs etc etc.
(edit: nvidia, not amd/ati radeon)
Supposedly the whole Nvidia debacle is the reason why Apple hardware doesn't sport Nvidia anymore. Though you can get them to work sometimes with iGPU, its more safe to keep up with AMD.
I'm not sure about FreeSync vs Gsync though, and I suppose Moonlight doesn't work on AMD.
There has been more[1] than one[2] case where GPUs in MBPs stopped working because of bad solder joints. And ifixit has a guide about how to fix your mac by putting it in the oven[3] to reflow the solder joints.
1: https://www.macupgrades.co.uk/store/product_info.php?product...
2: https://www.apple.com/support/macbookpro-videoissues/
3: https://ifixit.org/blog/6882/why-i-drilled-holes-in-my-macbo...
Even for the eventual repair order, Apple stuck in the same logic board with the same parts. I think the main difference was that those logic boards throttled the CPU to run cooler.
That early 2011 Macbook Pro was the first laptop I ever had (by any manufacturer) that died. I was lucky I still had the MBP around by the time Apple issued the repair order. I was able to swap out my dead logic board and eventually unload the laptop.
If MacBooks are famous for it, so are all laptops with an nvidia gpu.
The PC vs Mac hardware dichotomy is best summed up as, "Apple gives you less choice so that they can exercise more quality control."
Dell sells way more PCs (in much more varied configurations) than Apple does. They're generally more repairable, and they're generally cheaper. They're also willing to let you max out specs and build weird configurations; configurations that arguably don't always make much sense. That range means they also sell a few objectively cruddy laptops, so you need to be careful and read reviews and do comparisons before you buy.
Apple positions itself as different from those companies. The comparison I'm making above isn't something that only self-building PC nerds understand; the same explanation is what I give to regular consumers who are trying to decide whether to go with Windows or Mac - people who will never try to build their own stuff.
"Should I buy a Windows PC or a Mac?"
"Do you want to spend a lot of time researching brands, reading reviews, and thinking about hardware specs?"
I think the point still stands that Apple is positioning itself in a different category than companies like Dell. If not, I'm gonna have some complaints about that $3K price tag.
The closest comparison is something like the Microsoft Surface - an expensive, premium device, with low repairability and configuration options, but (theoretically) increased reliability with a streamlined experience that "just works" out of the box. When the Surface line has problems, I put more blame on Microsoft than I would put on a company like Dell in an equivalent situation.
You've just made this up and now are criticising Apple/Microsoft for it.
I wasn't criticizing Apple or Microsoft for anything when I said their laptops weren't repairable, I was giving them the benefit of the doubt that some of their engineering decisions around repairability are made for good reasons.
I don't think that's a particularly controversial idea. I mean, no one is seriously going to argue that they are repairable. So assuming there's a good reason for that, this is the standard explanation that I and other people would give to someone complaining about repairability - that Apple optimizes for other things.
Heck, it's the explanation that Apple gives: "Our devices are more reliable when we're the only people messing with their internals."
More than one case in 20+ millions units sold?
I'm not sure that qualifies as "famous for".
Have you noticed how low the bar is for "famous" these days?
Whether it's some FaceTube "star" or tech "scandal" du jour, everything is massive these days and everyone on the planet is affected. Somehow, there are no small issues anymore.
Is it saying "don't go robbing the masses"? Is it agreeing with the parent's comment by means of a literary reference that I thought was an approachable and analogous representation of the phenomenon the parent was describing? Seems whoever flagged this read it in a wildly different way than I intended it. Would it not be "this" if I had worded the same point differently--e.g., "You know, you're right, parent. These events totally remind me of 1984's Two Minutes Hate."?
Genuinely curious what "this" is that I shouldn't do, dang, because I've spent 8 years intentionally avoiding doing it.
The issue isn't just the quality of a comment in its own right—it's the quality of what it leads to. These things compound.
Thanks for being a fine HN commenter and user!
There's another thermal control unit involved and what people found was that the firmware for it was causing the CPU to drop down to 800mhz instantly when it detected the specific terminal event and when it reset that, it went back up into high-overheating load (such as the benchmark sustaining the high load) instead of staying at the right balance. In other words, it went into a loop of super low power load and then super-high load instead of the middle where the terminal balance is right for the load it is doing. Because of this low/high loop, the performance was not consistent and degraded as the overheating cause it to drop to 800mhz all the time.
Where Apple likely messed up is that the firmware was not placing the CPU at the middle to avoid going into a loop, this allows the CPU to sustain a better CPU performance consistently and for an extended period of time.
Apparently the power configuration is the one Apple's always used but previous CPUs didn't come close to the current generation in ability to draw power, so they'd never reached the VRM's thermal limits.
By lowering the CPU's maximum TDPup to 50~55W, they avoid the VRM over-heating and while the CPU does not turbo as high as with a 100W TDP allowance it can stay there just fine.
Basically based on the location and type of sensor(s) and your cooling solution you will apply an offset to get the actual temperature to be used for managing the power profile and cooling of the component.
This is pretty common and the reason why many apps don’t readout the temperature of newly released components correctly.
Ofc it can be something completely different and Apple is hiding the actual cause but I don’t deal with tinfoil fashion accessories.
Which is a shame because the theme this year is I, Robot.
>> Originally posted at MacRumors by winterny >> https://forums.macrumors.com/threads/optimal-cpu-tuning-sett...
These "magic codes" are probably just #define macros somewhere in the kernel repo, in my case it was in [1]
[1] https://github.com/torvalds/linux/tree/master/drivers/platfo...
Link to his channel: https://www.youtube.com/channel/UCVYamHliCI9rw1tHR1xbkfw
ftfy
Having a chip which can max out thermal load while plugged in at your desk, with the expected thermal results, and then throttle down when you go mobile, again with the expected thermal results, is actually the best of both worlds.
And with one click the user can actually choose which power mode they want to operate in.
A laptop that throttles back to avoid getting hot when you need the CPU/GPU operating at peak performance would be much less useful, particularly for professional use.
What you're referring to is a branding tactic. It has nothing to do with some subconscious direction on how to use the device. It's not only reasonable, but expected, that customers will use the device in their laps... no matter what the naming convention. Come on.
Edit: It was a Dell Latitude that burned the man's penis: http://ftp.compal.com/Download/NB/NCL50/User-Service%20manua...
Oh, really? Let's see.
Dell: https://www.dell.com/en-us/shop/dell-laptops/sc/laptops
HP: https://store.hp.com/us/en/cat/laptops
ASUS: https://www.asus.com/us/Laptops/
Lenovo: https://www.lenovo.com/us/en/laptops/c/LAPTOPS
Toshiba: http://laptops.toshiba.com/computers-tablets/home/
Acer: https://us-store.acer.com/laptops
Samsung: https://www.samsung.com/us/computing/windows-laptops/
I will stop there, because it's obvious that 'the entire industry' did not make a 'huge hard stop and turn on the terminology'. Stop being a fool.
If that's not the case anymore, then that's just because I haven't bothered to keep up with the industry as I no longer work in comms for PC manufacturers. As I said, I haven't put any thought into it recently. I'm neither being a fool nor am I lying, as you seem to be implying. Times change and I prefaced my statement to admit that the info might be outdated.
I can't even believe you put all that effort into collecting those links. That literally does not change the fact that there was an industry-wide aversion to using the term "laptop".
I guess that's normal for someone fixing Apple stuff as his day job and thus always only seeing apples bad side. Especially with seemingly Apple not wanting such repairs to happen.
My 2008 alu worked without a hitch for 10 years until a bad third party battery started to give problems together with the keyboard a month ago and I decided not to try to fix it anymore: that wasn't a bad investment.
It would be interesting to get some perspective from a longtime Apple employee on this.
I doubt it. They've had problems for years... bulging batteries, bad solder, bendy iPhones, etc.
But, is the quality really going down, or is Apple just an easy target? I tend to think it's more of the later. I've had Lenovos and Dells for work and they've not been perfect.
I have installed the update and so far so good. I have only had this 2018 model for a little over a day but I am loving the keyboard changes and TrueTone display and performance for me is superb.
1: https://www.theverge.com/circuitbreaker/2018/7/19/17590100/a...
https://www.lenovo.com/us/en/tablets/android-tablets/yoga-bo...
I reckon keyboards are a matter of taste ultimately.
People can debate about the feel of the keyboard but its fact that the new keyboards are louder compared to the 2015 keyboard. The noise gets annoying when you are working late night.
Even the noise is a subjective taste, ultimately.
TL;DR The CPU is stuck in a cycle of switching between minimum clock frequency and turbo speed.
When I got my i9 mbp I noticed the fan was quite active and it was fast, but not significantly better than my 2014 mbp quad core. I just updated with the fix this morning and so far I notice way less fan usage and it did seem to be a bit snappier. I will run some benchmarks later and see where it lands. If it isn’t working I’m about ready to abandon platform. Sucks because my entire music studio is Mac based.
I expect the next Mojave beta release to resolve this, but don't expect the next release to come faster due to this bug.
Temps still high, up to 95% during cb but I guess that's expected with this type of laptop body under extreme load and can be somewhat mitigated with external fans.
I was going to return this but will keep it now.
I hope it is though. I was getting really depressed at the idea of switching to something else.
The Macbook Pro from 2008 literally had a thermal issue. It made some units not boot at all. And it required a logic board replace, not just a software fix.
https://www.ifixit.com/Wiki/MacBook_Pro_15%22_Unibody_Troubl...
"If the fans run at high RPMs, but the computer does not boot, there is probably an issue with the thermal sensors.
Late 2008/Early 2009 models feature a thermal sensor on the heat sink. First, ensure that the thermal sensor on the heat sink is plugged into the logic board. If it is, try replacing the heat sink. If this does not fix the problem, the logic board most likely needs to be replaced."
Just running an already hot laptop hotter can't be good.
And the system as a whole will still run quite hot for high CPU+GPU loads, so who knows how that affects the electronics and battery life longer term.
Please, Apple, just make it 1mm thicker already.
[0] https://www.macrumors.com/2018/07/13/2018-macbook-pro-fastes...
Apparently, the OS does this to prevent thermal damage. It perceives that the CPU is overheating, so it sends one core into an idle loop to prevent it from doing real work.
Finally, I took it to an Apple repair shop whose diagnostics determined the CPU thermal sensor was not working. That didn’t immediately help. After much web searching, I found that a fix for this it was to completely disable the power management profile by deleting it. Now the machine works perfectly normally in terms of everything - heat, fans and performance. The machine seems to work identically to its original behavior, so I am puzzled at what function the power management profile performs.
So now it just don't give a shit and fires away at the highest clock, without a care in the world about temperatures?
The clock speed should go down under "heavy thermal load", that's not a bug...
If the CPU draws more current than the VRM can handle, this is usually okay within bounds for a very short time, after than you'll get dropping voltage when the VRM starts to self regulate and the CPU will downclock in response.