Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Its clear, if you want an 2018 15" Mbp that is an real upgrade, you have to come from a pre 2016 models and buy the i7 base or at max, the 2.6 ghz, and to keep thermals as low as possible maybe even the 555X instead of 560X now that has 4 gb vRam but lower speeds.
 
OP TL;DR: Poor Apple didn't know any better! Dropping a 100+W power hungry piece of silicon into it's current 50-W cooling system....???

Just looked at the Intel's technical documentation. They say that 45W is enough to cool down the i9. Furthermore, from Intels 8-th gen data sheet Volume 1 (Table 5-1) defines TDP as following:

The average power dissipation and junction temperature operating condition limit, specified in Table 5-2Table 5-5 for the SKU Segment and Configuration, for which the processor is validated during manufacturing when executing an associated Intel-specified high-complexity workload at the processor IA core frequency corresponding to the configuration and SKU.

Sounds quite clear to me: when running "high-complexity workloads", 45Watt of cooling should be sufficient to maintain the base frequency. And MBP's cooling is good enough for that, as shown by the performance with earlier gen CPUs. Yet it doesn't for some reason. Also, there is no mention of 100W or anything like that anywhere in the documentation. And yeah, Apple should have tested nada, nada. Of course they did test. But don't you think its strange that the technical documentation of a SKU and its actual thermal performance are so different?

The i9 is designed to replace earlier top-end CPUs such as i7-6920HQ/7920HQ etc. With Coffee Lake, Intel does not offer a high-end i7 CPU anymore at all. And their top-end is as it seems clocked waaay too aggressively for the silicon to handle it.

My point is: either Intel is being a bit dishonest in their technical documentation, or we have some other issues that are not yet properly studied. Of course Apple carries responsibility. But I don't think its reasonable to tell that they shouldn't have used a chip that is marketed and described by its manufacturer as being suitable for their laptops.
 
Last edited:
  • Like
Reactions: mr.anthonyramos
Also, there is no mention of 100W or anything like that anywhere in the documentation. And yeah, Apple should have tested nada, nada. Of course they did test. But don't you think its strange that the technical documentation of a SKU and its actual thermal performance are so different?

When you're marketing and positioning a product as a "Professional" solution, with millions if not billions already spent in R&D over the years, this is inexcusable. You're essentially shifting the blame to Intel because their documentation was misleading at best?

It's Apple and Apple's fault only, since it is their OWN product. They're either incompetent and forgone simple run-of-the-mill testing, the results of which couldn't be any clearer: can't cool a 100+W CPU with a 50-W cooling solution (Source: 100W), or they slabbed it in anyway - which is a hallmark of arrogance and disdain towards its customers.

It's basic trivial testing we're talking about; and for a company with $ billions stashed off-shores, this saga is pathetic and has deservedly tarnished their reputation and overall image even further. Just sad. There is simply no excuse.
 
Probably 60% Intel for making and offering these rather dubious ‘45W’ chips and 40% Apple for using them anyway...
 
and for a company with $ billions stashed off-shores, this saga is pathetic and has deservedly tarnished their reputation and overall image even further. Just sad. There is simply no excuse.

And not offering the i9 wound't tarnish their reputation, when their competitors use it in similar laptops (which don't have better cooling either)? What you say is very reasonable, but I feel like you don't really look at the bigger picture.

Anyway, I am looking froward to testing out my i9 MBP, hopefully arriving next week. The throttling reports I saw so far strike me as very dubious. I want to look at it myself before formulating a final opinion.
 
Take the 8550u.

It’s rated at 28w TDP, but can be adjusted as low as 15w.

Razer was wise enough to know that their Blade Stealth could not cool the 8550u in its current chassis. Therefore, they dialed it in at 15w to avoid thermal throttling.

Dell uses the same 8550u and pegged it at 28w. Guess what, it throttles.

Lenovo dialed in their 8550u at 28w and it throttles the worst of the bunch.

Manufacturers have a choice and range to work with when dealing with these chips. However, thave to pay attention and choose to do so
 
Say what you want, but if a machine can't even sustain the base clocks under sustained load that's definitely a design defect and it should absolutely come up during standard pre-release testing. Thus it's clear beyond any reasonable doubt that Apple knew about this and should have either beefed up the cooling of the i9 machines the same way they water cooled the higher end PowerMac G5s or then just not released the i9 machines. Instead they chose to do neither, which is quite honestly an absolute disgrace.

The fact that sustained load benchmarks show the new i9 machine actually losing to the higher end i7 machines from last year's lineup is simply inexcusable. Only way it could be excusable is if this year's models were cheaper than last year's, but they just aren't.

But yeah... Being too design driven has long been a complaint regularly leveled at Apple and I believe that over the last few years it's been more warranted than ever before.
 
What a dumb thread... How is there 3 pages?? Apple built the machine, not Intel; and Apple chooses what parts to use or NOT use. They design the chassis and cooling solution. Intel just builds chips. The end.

And by the way guys; we're still within two weeks of release - so everyone here who takes issue with the i9 situation can return/replace it. So the outrage from some is a little weird to me. I have i9 on order and plan to return/exchange for something else at pick-up. Not that worried. Apple can figure out what to do with all these returned i9s...
 
  • Like
Reactions: The Mercurian
Probably 60% Intel for making and offering these rather dubious ‘45W’ chips and 40% Apple for using them anyway...
Intel for offering?? yeah, tell me a company thats says NO to money and profit
Or give me an official document that shows us Apple was obligate to buy these and use these i9
 
Say what you want, but if a machine can't even sustain the base clocks under sustained load that's definitely a design defect
I haven't put my MBP under load yet, and it was definitely spiking in the 90s, during my initial setup, but with Macs Fan Control the temps are back to what I consider normal. I'm hoping that app will fix the issues for me.

I think computer manufacturers were caught flat footed with Coffee Lake. I think many of them including Apple just assumed they could plop the chip set into the existing enclosures and move on, not so. Coffee Lake requires more aggressive cooling. Razer Blade did this and I believe MSI is another where they customized the cooling solution and both run cooler (relatively speaking)
 
Probably 60% Intel for making and offering these rather dubious ‘45W’ chips and 40% Apple for using them anyway...
Large computer companies like Apple get early advanced copies of the chipsets. I think many of them just sat on their laurels and didn't really do the due diligence, planning and work needed - just my $.02
 
Intel for offering?? yeah, tell me a company thats says NO to money and profit

In a situation where there is only one noteworthy chip supplier and all of your competitors are adopting their chips (which have heat issues in most laptops btw), I don't think you can afford not to adopt them. A loss of reputation for Apple from having a single generation of machines with situational throttling issue is probably less than the loss of reputation from not updating to the 6 cores CPUs.
[doublepost=1532085006][/doublepost]
Razer Blade did this and I believe MSI is another where they customized the cooling solution and both run cooler (relatively speaking)

Razer Blade never had exceptionally good CPU cooling (at least compared to more "serious" gaming laptops). From what I've seen, the 8-th gen in it does suffer from a good deal of thermal throttling.
 
Intel don't deliver the thermal performance, Apple don't provide a case that accounts for this higher than stated thermal performance - even though Apple has years of Intel not meeting thermal targets. Everything throttles and it isn't Apple specific, the Dells throttle, Asus, HP - everyone! The worst part is the Apple tax on the hardware to get the same limitation as the other manufacturers.

Apple need to release Macs with A Series chips. It's getting to the point that when you look at the iPad Pro 10,000 geekbench versus the Macbook 12'' 5000 geekbench it would be worth taking the performance hit in software emulation of X86 - like Rosetta used to in the PPC to Intel transition. Imagine an A12x in the case of a 13'' Pro, plenty of room for cooling.
 
IFrom what I've seen, the 8-th gen in it does suffer from a good deal of thermal throttling.
No it doesn't. They moved to a vaper chamber and its running a heck of lot cooler, especially compared to the MBP. I know, becauseI bought one. I also did a lot of research and under heavy load, it was hitting the 80c. This is even more remarkable given the high end GPU that's in the Razer which generates a lot of heat on its own, where as the MBP has a rather anemic GPU.

I opted to return the Razer, not for thermals, but I wanted to see what Apple did, and I was pleased to see a 3rd gen keyboard. I now have the MBP in my hands, and comparing the two, MBP and Razer, early signs point to the MBP running a hotter then the Razer.
 
They moved to a vaper chamber and its running a heck of lot cooler, especially compared to the MBP. I know, becauseI bought one.

Interesting to know, thanks, was not aware of that! The reason why I was saying that it throttles is because it seems to perform below the Dell XPS 15" with the same CPU (https://www.notebookcheck.net/Razer-Blade-15-i7-8750H-GTX-1070-Max-Q-FHD-Laptop-Review.305426.0.html vs. https://www.notebookcheck.net/The-X...wice-as-fast-as-its-predecessor.308578.0.html).
 
  • Like
Reactions: The Mercurian
Apple could have simply refused to implement these processors into their product.

If Apple would have posted a statement claiming the chips were not within their spec/thermal guidelines, it would have been just that.

Apple knew the thermals of these chips long before we did.

If they were wanting to stick with the same chassis and knew in test labs that it was going to be an issue, they did not have to release the product.

Just because the market is pressuring them to release a product does not mean they have to follow. (That basically Apples additude to begin with) “Think Different”

If they would have designed the chassis to accommodate the CPU, this would not have been an issue.

Honestly, I think this more or less fuels their desire to move to their own architecture.

If they can fool the public into thinking that Intel chips suck, then they won’t face as much pushback when the transition is made.

I've thought the same for some time, and there's some logic to it with Tim and Co once again procrastinating about how awesome the 2020 MBP is, how much better than all other MBP's. Totally blanking the PC market as if it never existed.

I don't think for one second Apple is clueless and has no idea what it's doing. Setting the stage for expensive hardware designed for obsolescence with ever more generous margins.

Apple just wants the illustration of "Pro" to make it's target audience feel like a Pro and feel better about their expensive purchases...

Q-6
 
Interesting to know, thanks, was not aware of that! The reason why I was saying that it throttles is because it seems to perform below the Dell XPS 15" with the same CPU (https://www.notebookcheck.net/Razer-Blade-15-i7-8750H-GTX-1070-Max-Q-FHD-Laptop-Review.305426.0.html vs. https://www.notebookcheck.net/The-X...wice-as-fast-as-its-predecessor.308578.0.html).
All the benchmarks I saw, were pretty fast, I never heard about notebookcheck, but from my research they had the best Coffee lake/GPU combination that ran much cooler then almost anyone else. I think MSI was cooler
 
I haven't put my MBP under load yet, and it was definitely spiking in the 90s, during my initial setup, but with Macs Fan Control the temps are back to what I consider normal. I'm hoping that app will fix the issues for me.

I think computer manufacturers were caught flat footed with Coffee Lake. I think many of them including Apple just assumed they could plop the chip set into the existing enclosures and move on, not so. Coffee Lake requires more aggressive cooling. Razer Blade did this and I believe MSI is another where they customized the cooling solution and both run cooler (relatively speaking)

I tend to disagree, within the first hour it was clear to me that the 8750H was almost a doubling it's power draw under full Turbo.

Intel is provides engineering samples way in advance, some companies did their homework some did not, equally none were surprised.

Apple simply doesn't care or doesn't want to redesign the chassis, preferring It's customer's to pay more and suffer worse performance.

Q-6
 
Intel is provides engineering samples way in advance, some companies did their homework some did not, equally none were surprised.
Yes, that's what I'm saying, but I said flat footed because I think Apple and Dell both did one of two things.
1. Didn't test that much and just plopped them into the existing designs.
2. Did test and revealed the throttling issues, but chose to release them anyways.

In either case, its the manufacturer's fault and not Intel's. Like I said, other computer makers like MSI and Razer found cooling solutions for the hot Coffee Lake CPU.
 
  • Like
Reactions: throAU and Queen6
Like I said, other computer makers like MSI and Razer found cooling solutions for the hot Coffee Lake CPU.

I'd say that their cooling solution was aimed at cooling down the large GPUs they use. I very much doubt that they did it for Coffee Lake.
 
I found both guilty. Intel has became complete liar to me. Look at X series chips, i7-8086k parody, meltdown & spectre, complete false tdp values for 8th gen cpus, and 28 core cpu fiasco..
Apple on the other hand should have redesign the cooler.
well.. guys we are having all theseeeeeee problems because Apple made the case 3 mm thinner and 200 lighter... bravo!
 
  • Like
Reactions: matram and Queen6
Just looked at the Intel's technical documentation. They say that 45W is enough to cool down the i9. Furthermore, from Intels 8-th gen data sheet Volume 1 (Table 5-1) defines TDP as following:



Sounds quite clear to me: when running "high-complexity workloads", 45Watt of cooling should be sufficient to maintain the base frequency. And MBP's cooling is good enough for that, as shown by the performance with earlier gen CPUs. Yet it doesn't for some reason. Also, there is no mention of 100W or anything like that anywhere in the documentation. And yeah, Apple should have tested nada, nada. Of course they did test. But don't you think its strange that the technical documentation of a SKU and its actual thermal performance are so different?

The i9 is designed to replace earlier top-end CPUs such as i7-6920HQ/7920HQ etc. With Coffee Lake, Intel does not offer a high-end i7 CPU anymore at all. And their top-end is as it seems clocked waaay too aggressively for the silicon to handle it.

My point is: either Intel is being a bit dishonest in their technical documentation, or we have some other issues that are not yet properly studied. Of course Apple carries responsibility. But I don't think its reasonable to tell that they shouldn't have used a chip that is marketed and described by its manufacturer as being suitable for their laptops.

Very well said!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.