Tests Confirm Apple's Throttling Fix Improves Performance for 2018 MacBook Pro Models [Updated]

[IG88]

Here is my problem with this: There is a lack of QA/QC that used to be present for all Apple products before they were ever sent out. Something this simple should have been caught before the first unit was shipped.

Agreed.

 

[Altis]

Yup, absolutely. Kind of why this whole thing has been a bit overblown in my eyes, but I can understand not getting advertised performance. Good on Apple for fixing the issue within a week of release.

If only Apple could be bothered to test this stuff rather than leave it up to reviewers to do their legwork.

 

[DNichter]

If only Apple could be bothered to test this stuff rather than leave it up to reviewers to do their legwork.

That's fair. They very well might have and were further along in their release cycle and decided to release a fix in the first week.

 

[justperry]

My guess is this is a reference to a plist type dictionary and not an encryption key. In other words the “key” was a missing key-value pair that defined a preference for how power management worked for the new systems. With that key missing, the power control management system dropped back to some default behavior which resulted in the throttling.

A key (in a few kb plist file) of over 1 GB

I guess though Apple released a more than 1 GB update just to make it (very) hard to search for what really changed.

I do think Apple is being honest, they just messed up and they provided a fast patch which works.

 

[IG88]

Then why do all these macrumor ppl and you tubers claim it was a technical glitch ? Not that I don’t believe you but that would make all these folks go on my blacklist. Like the guy above you says the issues is fixed - no it’s not fixed it’s patched with a bandaid.

I don't think it was a glitch. I wouldn't call it a band aid fix either though.

For example, 2017 non touchbar 13" MBP on notebook check.

https://www.notebookcheck.net/Apple...017-i5-without-Touch-Bar-Review.234282.0.html

Scroll down to the bottom under Energy Management. Look at the Power consumption, Load Maximum compared to the Dell XPS 13. Obviously Dell put a much lower peak power limit on the CPU than Apple did.

As I said, I believe that Apple has historically & intentionally put in a big power limit to just let the CPU throttle on temperature. There isn't anything sinister about it. Do I think it's a great idea? Not really but it's not horrible either.

I think these 6 core CPUs kinda bit them in the ass on that and they now have to put in a power limit so it doesn't overload the VRM in peak power demand. It isn't necessarily a bad thing, they probably should have been putting in reasonable power limits all along.

Upsizing the VRM wouldn't likely change much really, you can't pull more power in the long term than the CPU cooler can dissipate.

 

[avkills]

Give Apple a break, I mean they are probably too busy testing the Mac Mini Pro that legos together via some unknown port to be bothered testing a laptop design that has already been tested.

I mean why pay for on staff hardware testers when you get that for free with pimpled faced kids living in their parent's basements running YouTube shows.

 

[curtvaughan]

It is a Windows machine. Apple supports Windows on it. It's just not a Linux, FreeBSD, etc machine.

I'll bet a fix in the Boot Camp drivers is coming soon. Obviously macOS is going to be top priority for them.

If I had the luxury of purchasing one of these, I'd try to install a double boot with Linux on it just to see how the laptop performs on the Linux platform. Assuming one could get Linux to install (turn off T2 "secure boot" feature), I'm guessing Linux and Windows would have similar performance issues on the new MBP. Might try to see if one of these youtube testers would be interested in trying to install and test a Linux distro.

 

[jonnysods]

Back in business!!

 

[Petetastic]

Had a few moments to mess around with the i7 2.6 last night.

I started the patch as soon as it turned on. While this doesn’t have to do with throddling, I expected it to generate a lot of heat as it updated. It was very cool to the touch. My mid-2010 would sear a steak to medium rare as it updates.

That’s good to hear. My mid 2010 2.6ghz is also an oven. I was really worried about this one being hotter.

Unfortunately, I ordered a 2.2ghz and it’s “preparing to ship” (past a point I can cancel). I may be convinced to return it and upgrade. I guess I’ll wait for more tests and see.

 

[VictorTango777]

Much improved. I see where the larger laptops from other companies may have better scores. But I mean, if you need a light machine that can do some moderate level of computing, this thing is a good buy. I mean, I'd like to see it 300-450 cheaper, but that's just me being poor. lol.

The 2016 and newer 15 inch MBPs are $300 - $400 more than the models they replace. Instead of $1999/$2499 we now have $2399/$2799 and still the same base RAM and storage capacity as before. I thought computer technology was supposed to cost less over time, not more.

To be honest I would kill to know what goes on behind the scenes at Apple. All of their resources and brilliant engineers, someone MUST’VE seen this whole thing becoming a future issue.

Like you said, all of the testing they do and they didn’t catch this? It’s kind of wild. I really wonder what the discussions looked like behind the scenes and if anyone was blamed.

On the other hand, did an engineer see an issue and try to say something but was silenced?

 

[OC40]

That’s good to hear. My mid 2010 2.6ghz is also an oven. I was really worried about this one being hotter.

Unfortunately, I ordered a 2.2ghz and it’s “preparing to ship” (past a point I can cancel). I may be convinced to return it and upgrade. I guess I’ll wait for more tests and see.

I should be able to get into the 2.6 tonight after I transfers over files. Im looking forward to it.

The 2.2 is receiving pretty good reviews. You may be able to walk into a store to exchange it by the time that you receive it.

 

[pipis2010]

Wouldn't it be nice to get that kind of Pro support for us regular plebs. All you need is a viral video and nearly 1.5 million YouTube subscribers.

These people are the influencers; are you an influencer? Maybe, but you can’t theoretically influence a million people, so no...

 

[curtvaughan]

Much improved. I see where the larger laptops from other companies may have better scores. But I mean, if you need a light machine that can do some moderate level of computing, this thing is a good buy. I mean, I'd like to see it 300-450 cheaper, but that's just me being poor. lol.

I would think the better choice for a light laptop for moderate computing would be something with the i7 or even i5 rather than the i9. It would fetch you the same form factor, perform moderate computing chores admirably, have less throttling issues, and be considerably cheaper.

 

[mrex]

People think they have to overspend on these things, but the reality is they don't. Now my USB-C hub was admittedly slightly more than your (around $30) because I also wanted: Ethernet, HDMI, USB-C pass through for charging, an SD card reader and a VGA/DVI port so that I can run two monitors off of one port. Really, all of that with using a single USB-C port and no additional drivers are required.

you can buy cheap devices or expensive, and the overall operation may differ alot. you can buy a 2m long ”tb3 cable”, a cheap tb3 cable for 30$ or an expensive tb3 cable for 60$. the difference is that one is 20gb/s and the other is active 40gb/s.

same goes with usb-c hubs. you can buy a hub with 20$ or 200$. and im not saying that you may not find a decent cheap alternative suitable for you but usually something is missing or altered especially when several ports are used at the same time. so it is not only that ”people pay to too much because they dont know that they could...”

 

[pipis2010]

The people who bought those i9s should thank the YouTubers. Fair play to Apple for fixing it quick, but one wonders if they would have bothered if there wasn't as much negative press.

I think they would have eventually fixed it. It’s a bit disheartening though, that Apple is waiting for public outcries, sometimes, before fixing things in advance...

 

[VictorTango777]

you can buy cheap devices or expensive, and the overall operation may differ alot. you can buy a 2m long ”tb3 cable”, a cheap tb3 cable for 30$ or an expensive tb3 cable for 60$. the difference is that one is 20gb/s and the other is active 40gb/s.

same goes with usb-c hubs. you can buy a hub with 20$ or 200$. and im not saying that you may not find a decent cheap alternative suitable for you but usually something is missing or altered especially when several ports are used at the same time. so it is not only that ”people pay to too much because they dont know that they could...”

There are times when you do "get what you pay for". And for the other times, there is Monster Cable.

 

[barmann]

This still won't shut the Apple haters up.

That's the more knowledgeable Apple users you want to shut up, how orange .

Done : Apple got a laptop working they delivered in a poor state - impressive .

To be done : Magsafe, keyboard fix, more varied and greater number of ports across the line, touchbar models must lose preferential specs and touchbar needs to become fully optional .

Non-soldered harddrives, user upgradable RAM and battery, ditch T2 chips .

 

[ThatGuyInLa]

Here's a question for you:

If you're on a plane without outlets, would you rather have the MacBook Pro with Retina and 2K resolutions to watch 1 HD video (2 hours) utilizing the CPU 100%, or a cheap windows laptop in SD to watch 4 videos (8 hours)?

For me, I'd prefer 8 hours of low resolution content, than 2 hours of high resolution content and 6 hours of nothing.

So, I think in the case of CPU throttling, each situation is unique to what you expect from your machine.

In your example, a BMW isn't much more luxurious than driving a Kia to a grocery store that is 5 minutes away. The luxury of a BMW only becomes apparent when you commute often for long distances.

I don't fly. I'm too fat. Seriously though, yes your scenario of course makes sense. It's why I use a $50 Kindle to watch TV in bed instead of an iPad. This way when I roll over on it and crush it or spill grape juice mixed with vodka all over it, it's not that big of a deal.

 

[curtvaughan]

Guys!! We only got a patch here, we didn't get the root issue addressed!

Yes! Apple fixed the bug it had and they do deserve a high five on that! Finding it, correcting it and within very short time roll a patch. All good and hard work! Yes, clap our hands for that!

Now lets get back to the root issue here as David Lee points out ALL of the thinner systems fall victim to the lack of adequate cooling so the i9 is throttled by the systems design.

While the constraints may be livable for many, there is still a very large crowd that want and need more from this chip! Apple failed them! This chip could still give us 10 to 20% still better performance!

The answer to this issue, at least with current chip technology, is that for the highest performance in high powered CPUs / GPUs to be accessible there must be powerful cooling supplied, and that currently requires thicker, heavier, fan laden laptops, or high performance destop workstations. That might not be the case in the future, if chip designers can come up with cooler processor technologies and more efficient ways to dissipate heat generated under compute-intensive processing. That several competing "thin" i9/6 core laptops are experiencing nearly identical problems seems to indicate that thin and light for high performance isn't ready for prime time quite yet.

 

[Petetastic]

It's why I use a $50 Kindle to watch TV in bed instead of an iPad. This way when I roll over on it and crush it or spill grape juice mixed with vodka all over it, it's not that big of a deal.

*spit take*

 

[Spectrum]

Based on the videos I've seen, I'm still wondering if the advertised turbo speeds can reached on these laptops?

 

[Shanghaichica]

I repeat. Do you have intimate knowledge about that? You are just wildly guessing. Apple said they did work with David Lee. Why would you doubt Apple all of a sudden?. Because it fits your Dave Lee bashing theory better?

I never disputed that they worked together, just the purpose of them working together.

 

[fathergll]

I don't think companies changed focus they have only reacted to the purchasing trends of companies and individuals as a whole. Staying in a company/corporate mindset. In the past when laptops/desktops were much bigger dedicated, heavy duty server hardware was always used to perform heavy tasks. Now with amazon and other cloud computing platforms server farm capacity can be scaled up rapidly to provide computing resources as needed. Companies will continue to focus less on desktops/workstations as more and more applications and computing resources can be served by on demand cloud solutions.

Correct, companies follow the money but even if cloud solutions will continue to grow the simple fact is they are selling expensive hardware aimed at doing heavy tasks local to the users. Why sell a computer that starts at well over $2k if the user is supposed to use a cloud?

 

[curtvaughan]

Give Apple a break, I mean they are probably too busy testing the Mac Mini Pro that legos together via some unknown port to be bothered testing a laptop design that has already been tested.

I mean why pay for on staff hardware testers when you get that for free with pimpled faced kids living in their parent's basements running YouTube shows.

I agree with your sentiment, but I failed to notice any of the youtube MBP testers of prominence this last week sporting excessive pimples or hanging out in basements. Most seemed to be casually dressed and competent with computer hardware issues. Some of these guys, like Louis Rossmann and David Lee, appear to be quite professional. I'd trust Rossmann with any of my computer repair issues (PC or MAC), having witnessed him tear down and repair computers online. Apple obviously respected Lee's testing and commentary.

 

[falainber]

No, Dell's XPS suffer from the same issue. No update has been released to address this.
https://www.ultrabookreview.com/14875-fix-throttling-xps-15/
https://www.dell.com/community/XPS/Dell-XPS-15-9570-i9-thermals-throttling-and-solutions/m-p/6116619
https://www.dell.com/community/XPS/XPS-9560-Repasted-and-undervolting-still-throttling/td-p/6065052
http://forum.notebookreview.com/threads/dell-xps-15-9570-benchmarks-temps.817970/

You seem to be under impression that this fix prevents throttling. It does not. It improves management. In cases that you referenced Dell laptop was running different load from what MBP was subjected to. In the similar tests, as per this article, Dell behaved as good as "fixed" MBP. Gaming laptops did better. The problem for Apple customers here is as follows. XPS is a consumer/lifestyle model. In addition, Dell have gaming laptops (Alienware) and Pro laptops/workstations (Mobile Precision) which have different form factors and much better cooling systems. Apple on the other hand sells only lifestyle models pretending they are professional devices.