The MBP throttling issue isn't what everybody thinks...

[PeterBocan]

1 minute ago, Shorty88jr said:

I agree software on point hardware the most horrible thing ever designed it feels like they hired monkeys to build their computers while the employees worked on software.

Well I think it's mostly due to consumerism that Apple introduced into the IT world. They have to prop up the sales from time to time and now as they are on the border of "unusable & overpriced" they are looking for some must-haves for plebs. 

[PlayStation 2]

The threads about the MacBook Pro refresh really remind me as to why I absolutely abhor blind hatred and blind fanboyism together.

I'll get the easier part out of the way, the latter. Many ardent Apple fans, even on this forum, have admitted that this MBP refresh is, well, not good, to say the very least. While they may tend to defend what could reasonably be bashed, this time around, everyone knows it's just a disaster all-around.

Now, the former. JE-FUCKING-SUS. I know, this laptop really shouldn't have been launched. Apple really should've put more thought into this refresh. But even where Apple deserves credit in areas (well, not this laptop obviously), these kinds of people would rather dick ride something else rather than give credit where credit's due.

 

I try not to fanboy. I call them when I see them, and it's why I stopped being an AMD fanboy. I realized that, well, AMD's CPU department wasn't where it needed to be back in 2014 and 2015. Apple can have both praise (their LCDs are probably some of the best screens in general on the market, and iOS optimization) and criticism (this laptop, the way iOS 8 launched, their lack of updates on the Mac Mini). I just don't see why people have to do one or the other.

[leadeater]

3 hours ago, abazigal said:

We have seen what it takes to create a laptop that takes full advantage of the i9 chip. No way in hell Apple is making a device that thick and with that poor a battery life. 

Apple might not want to make a laptop that thick, or thicker than currently, however doing so will not make it have poor battery life, quite the opposite in fact. More volume means more battery volume, more battery volume means more stored energy, more stored energy means longer device operation.

 

An i9 in an Alienware, Acer, Asus or MacBook Pro is still the same processor and it doesn't use more energy because it's not in a MacBook Pro or in a thicker laptop.

 

I'll grant you that in a laptop with better cooling it will be able to use more energy because it no longer has the handbrake on but you shouldn't drive around with the handbrake on, not a good idea.

 

The 0.1 inch and 200g saving is not worth it for the design and cooling challenges that comes with that, not once did I ever look at a 2015 MacBook Pro and think it was too thick or want it to be thinner. MacBook Air is a thing.

 

3 hours ago, abazigal said:

To me, both sides are guilty. Intel, for shipping an i9 chip with the specs that they did, and Apple for still trying to go ahead and cram it into their devices regardless. 

Did Intel require or force Apple to put the i9 in the MacBook Pro? You'll have the answer to whether or not Intel has any fault at all in this with the answer to that question.

[Guest]

TBH in my view it's still both Intel and Apple's fault. This does happen in other i9 notebooks. We don't really know whats going on behind the scenes but I'm thinking Intel marketed the i9 to Apple as a replacement for their top of the line i7, and in doing so did not discuss with Apple the power limitations. They probably thought they could just pop the i9 in. 

[Jito463]

As one who dislikes both Apple and Intel, I believe I can say I'm sufficiently unbiased in this issue (in terms of not taking sides towards one or the other).  In my opinion, Intel doesn't deserve any blame here.  Yes, the i9 is a hot running chip that draws a significant amount of power, but it's not their fault that Apple tried to cram it into a system which was never designed to handle it.  Apple's engineers are the only ones to blame here.  They should have redesigned the whole thing to conform to the cooling needs of the CPU, not tried to conform the CPU to the cooling capability of the laptop.

[PeterBocan]

15 minutes ago, RorzNZ said:

and in doing so did not discuss with Apple the power limitations

That's BS. The first thing you ask Intel: what's the TDP and power usage. 
 

 

[Guest]

12 minutes ago, PeterBocan said:

That's BS. The first thing you ask Intel: what's the TDP and power usage. 
 

Depends how things are marketed to the company. It wouldn't be too far to assume that because it's a mobile chip it wouldn't draw too much and would be relatively cool. 100W is not a standard at all for a mobile chip. 

[GilmourD]

3 hours ago, DrMacintosh said:

Intel can't make a cool CPU. 

I would never expect an i9, their performance part, to run cool. Maybe I've been doing this for way too damn long but in my head higher performance parts have a higher TDP and have always run hotter, requiring better cooling. It's been that way since they started shipping machines with HSFs (which is weird to think that my first two machines, an Intell 8088-2 and an AMD 386DX-40, didn't have any active cooling directly on the chips).

[Jito463]

2 minutes ago, RorzNZ said:

Depends how things are marketed to the company.

Marketing is irrelevant.  Unless the engineers were overruled by management (which can and does happen), then their engineers would have to be idiots not to get the actual numbers from Intel.  Marketing is for sales, raw numbers are what's needed for engineering.

[PeterBocan]

5 minutes ago, RorzNZ said:

Depends how things are marketed to the company. It wouldn't be too far to assume that because it's a mobile chip it wouldn't draw too much and would be relatively cool. 100W is not a standard at all for a mobile chip. 

Okay, let's assume Intel marketed to Apple Core i9 and somehow Apple's executive officer got so confused he never checked technical specification of the processor. Who's to blame for this blatant incompetence? Intel markets the TDP as they have defined it. Apple should have known better.

[D13H4RD]

I think it’s a bit of both, actually. The VRMs explain the drops to 800MHz and such (a la Dell XPS15) and the cooling system itself also is unable to keep a CPU known to run really hot even in desktop replacements in check.

 

You’ve got to be daft to just blame Intel. Yes, they made a chip that runs super hot and has crazy cTDP runaway at very high clocks. But doesn’t Apple, Dell and all the others have a choice of not using them? If Apple didn’t want to make a major redesign of the chassis and cooling solution to accommodate a toasty CPU, they probably shouldn’t have offered it in the first place.

 

Yeah, yeah. Marketing blah blah blah. You either make it work or you get egg on your face when it fails.

[Guest]

9 minutes ago, PeterBocan said:

Okay, let's assume Intel marketed to Apple Core i9 and somehow Apple's executive officer got so confused he never checked technical specification of the processor. Who's to blame for this blatant incompetence? Intel markets the TDP as they have defined it. Apple should have known better.

Marketing is about creating a point across. If you write down a spec sheet it's not up to the officer to read what you have written, it is about communication of that information. This will impact Intel's sales as well as Apple's. Both sides failed to do their jobs and as a result both products will not sell. 

[ZacoAttaco]

13 minutes ago, RorzNZ said:

Marketing is about creating a point across. If you write down a spec sheet it's not up to the officer to read what you have written, it is about communication of that information. This will impact Intel's sales as well as Apple's. Both sides failed to do their jobs and as a result both products will not sell. 

Well Intel already have a contract with Apple, they supply the chips and are compensated, so even though i9 model won't sell, Intel are still making money, while they are taking some sort of a financial hit, it's Apple who will see this affect their bottom line more so.

[Guest]

Just now, ZacoAttaco said:

Well Intel already have a contract with Apple, they supply the chips and are compensated, so even though i9 model won't sell, Intel are still making money, while they are taking some sort of a financial hit, it's Apple who will see this affect their bottom line more so.

Potential sales are still sales. Apple might not order more until this is resolved. Remember as well this isn't just an Apple only issue (but more prevalent and people complain about Apple a lot more so in the media etc). 

[HammerDrop]

https://ark.intel.com/compare/134903,134906,134906,134899

 

all 45w tdp 

 

no marketing, no bs, just specs to design by. If you tell me each chip has the same thermal requirements, then why would I design a different cooling mechanism for one and not the other?

[ZacoAttaco]

3 minutes ago, RorzNZ said:

Potential sales are still sales. Apple might not order more until this is resolved. Remember as well this isn't just an Apple only issue (but more prevalent and people complain about Apple a lot more so in the media etc). 

But are they going to resolve the issue though? They probably will just quietly stop ordering/selling i9 models.

[PeterBocan]

26 minutes ago, RorzNZ said:

Marketing is about creating a point across. If you write down a spec sheet it's not up to the officer to read what you have written, it is about communication of that information. This will impact Intel's sales as well as Apple's. Both sides failed to do their jobs and as a result both products will not sell. 

Man, you are going around Titanic with a duct tape trying to fix that 12 ft tall hole in a side. You blamed it on marketing now you are moving the goalpost to "communication". And in a few posts you will move the goalpost and claim that Apple Executives can't read. Stop defending the faults of the prime engineering from Cupertino, they don't need it, they don't deserve it. 
 

At the end of the day somebody had to read the specification, and even if nobody noticed a discrepancy, somebody had to work with the testing samples, they had to at least try it out. Or if they didn't try it out, then I would rather not buy another Apple device whatsoever. Your defence of the company is just rendering Apple as totally inept and incompetent as the computer manufacturer. 

 

8 minutes ago, RorzNZ said:

Apple might not order more until this is resolved. Remember as well this isn't just an Apple only issue (but more prevalent and people complain about Apple a lot more so in the media etc). 

Yes, Apple is under the scrutiny the most, because it's Apple that asks for 3000$ which is almost twice the price for Dell XPS 15". TWICE. Again... TWICE. You can't expect media to mourn on 1800$ where some issues can be expected. It's Apple that asks insane money and they don't deliver. 

[D13H4RD]

7 minutes ago, HammerDrop said:

https://ark.intel.com/compare/134903,134906,134906,134899

 

all 45w tdp 

 

no marketing, no bs, just specs to design by. If you tell me each chip has the same thermal requirements, then why would I design a different cooling mechanism for one and not the other?

45W of TDP at base clocks in heavy-but-realistic scenarios

 

That’s how Intel specifies their TDP ratings

 

It’s not unreasonable to expect Coffee Lake to have increased thermal demands due to more aggressive clocks and higher core counts.

[Drak3]

3 minutes ago, PeterBocan said:

Apple Executives can't read.

I thought everyone knew this.

[Guest]

3 minutes ago, PeterBocan said:

Man, you are going around Titanic with a duct tape trying to fix that 12 ft tall hole in a side. You blamed it on marketing now you are moving the goalpost to "communication". And in a few posts you will move the goalpost and claim that Apple Executives can't read. Stop defending the faults of the prime engineering from Cupertino, they don't need it, they don't deserve it. 
 

At the end of the day somebody had to read the specification, and even if nobody noticed a discrepancy, somebody had to work with the testing samples, they had to at least try it out. Or if they didn't try it out, then I would rather not buy another Apple device whatsoever. Your defence of the company is just rendering Apple as totally inept and incompetent as the computer manufacturer. 

 

Yes, Apple is under the scrutiny the most, because it's Apple that asks for 3000$ which is almost twice the price for Dell XPS 15". TWICE. Again... TWICE. You can't expect media to mourn on 1800$ where some issues can be expected. It's Apple that asks insane money and they don't deliver. 

Marketing is communication. I'm not defending the engineering, I never said it was fine, I said Intel and Apple are both to blame. 

[PeterBocan]

3 minutes ago, RorzNZ said:

Marketing is communication. I'm not defending the engineering, I never said it was fine, I said Intel and Apple are both to blame. 

Do you agree that Apple did not test new CPUs? 

[HammerDrop]

5 minutes ago, D13H4RD2L1V3 said:

45W of TDP at base clocks in heavy-but-realistic scenarios

 

That’s how Intel specifies their TDP ratings

 

It’s not unreasonable to expect Coffee Lake to have increased thermal demands due to more aggressive clocks and higher core counts.

Reports show that the the core i9 is unable to stay at its base clock but rather throttles down to 2.3ghz. So either number is wrong, or tdp is a bad measure of thermal requirements. Not arguing against either. What I'm arguing against is the claim that engineers did not design or account for the heat. You can only design based on the specs given to you. In this case the only spec given is 45w tdp... for both i7 and i9

 

it doesn't matter if it's "unreasonable to expect coffee lake to have increased thermal demands"... so don't give me a 45w tdp rating give me something higher 

[Guest]

8 minutes ago, PeterBocan said:

Do you agree that Apple did not test new CPUs? 

Not really, they probably did. Apple always do performance tests to compare to previous versions and advertise its xx% faster than the last generation. They only test applications like Photoshop etc, and not measuring clock speed and throttling, that would be for the hardware department to do (With these things blame gets passed around a lot - but they would have been tested certainly in that capacity).

[D13H4RD]

15 minutes ago, HammerDrop said:

 

it doesn't matter if it's "unreasonable to expect coffee lake to have increased thermal demands"... so don't give me a 45w tdp rating give me something higher 

That's the thing. You design thermal solutions to have more "capacity" than the rated TDP to account for times where the CPU will go over when boosted. 

 

Apple failed here. While the cooler may have been just barely adequate for a 45W 4C 8T Skylake/Kaby CPU, for the 8950HK, it's just a death sentence alongside the overloaded VRMs

[PeterBocan]

11 minutes ago, HammerDrop said:

Reports show that the the core i9 is unable to stay at its base clock but rather throttles down to 2.3ghz. So either number is wrong, or tdp is a bad measure of thermal requirements. Not arguing against either. What I'm arguing against is the claim that engineers did not design or account for the heat. You can only design based on the specs given to you. In this case the only spec given is 45w tdp... for both i7 and i9

 

it doesn't matter if it's "unreasonable to expect coffee lake to have increased thermal demands"... so don't give me a 45w tdp rating give me something higher 

So I looked up the 2017" Macbook Pro - They also had 45W TDP (with opt-in to 35W TDP) which are 4C/8T 
https://www.notebookcheck.net/Apple-MacBook-Pro-15-2017-2-8-GHz-555-Laptop-Review.230096.0.html
The current 8th gen has 6C/12T also in 45W TDP, which doesn't add up, as both are 14 nm Coffee Lake process.
https://ark.intel.com/compare/134903,134906,134899 

I can see that Intel maybe was not genuine BUT that doesn't excuse Apple for not checking and verifying. Maybe the Core i9 is actually like 60 W CPU. A monster.

4 minutes ago, RorzNZ said:

Apple always do performance tests to compare to previous versions and advertise its xx% faster than the last generation.

Apple can not claim that their brand new processor is XX % faster than previous if it's not cooled properly and there's throttling going on. That skews numbers and results, that renders all measurements void. You can not do that. You just can not claim something and not being sure that the CPU is cooled properly. Absolute sloppiness.