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

[Peskanova]

2 minutes ago, Morgan MLGman said:

This is a bit different case - ASUS Rampage/Maximus boards are over-engineered in terms of power-delivery cause they're designed for hardcore overclocking. They're designed to sustain ridiculous LN2 overclocks etc. Apple designed the motherboard themselves so they HAD to know and test the CPU & the power-delivery section and design it appropriately. It seems almost impossible that they didn't factor in temperatures in their testing so it really seems like they just wanted to make the MBP thinner and lighter while compromising on performance which doesn't seem smart considering that this particular laptop is aimed towards professionals who do heavy work.

This approach would only be acceptable with the MacBook Air in my opinion.

I dont think that puting better could afect the design, well put more can but I think that despite the solutions, Im afraid that this, instead fixing laptop cpu overthrottling issues (more or less gpu is fixed/better nowadays)  could set norm like notches or jack's on smartphones, if you are a brand and make an i9 laptop/ultrabook you now can say, well all laptops overthrottle, look the new mbp.... thats scary for laptop industry.

[leadeater]

7 minutes ago, PeterBocan said:

That's worth of additional 300$  

With an updated firmware and better fan curve and lower voltage I'm sure the i9 model can significantly improve from where it is at now and also consistently beat the i7 model, it still won't meet full potential but at least you wouldn't have wasted $300. So it's not impossible to turn that $300 in to actual performance gain, you may have to do it yourself though.

Edited July 24, 2018 by leadeater

[asus killer]

no engineer would allow this and it's ridiculous to even consider for a second they would do a laptop and not test it. The guys from marketing saw a i9 and wanted to used it and mark up the price of a macbook pro. 

I mean for guys in marketing it makes no sense to be able to sell something for x dollars more and not do it.

True not all marketing works like a potato powered brain but this happens a lot in real life, sometimes the surreal discussions happens in my company, it's just a matter of how much power they have in the company that they can get away in ignoring the engineers. I guess that in today's Apple the answer is a lot.

 

Hell even the motherboard manufacturers must noticed it in test phase. I believe Apple designs the mobo's but doesn't produce them.

[PeterBocan]

1 minute ago, leadeater said:

With an updated firmware and better fan curve and lower voltage I'm sure the i9 model and significantly improve from where it's at now and also consistently beat the i7 model, it still won't meet full potential but at least you wouldn't have wasted $300. So it's not impossible to turn that $300 is to actual performance gain, you may have to do it yourself though.

Paying for Lamborghini and have to tweak it by myself? HERESY! And praying that VRMs dont blow up on me. 

[leadeater]

Just now, PeterBocan said:

Paying for Lamborghini and have to tweak it by myself? HERESY! And praying that VRMs dont blow up on me. 

Must be the i9 Lamborghini:

[GoldenLag]

31 minutes ago, D13H4RD2L1V3 said:

Actually, some other laptops have the issue. 

 

Eurocom Q8, XPS15, AW15, etc. 

 

Many of them don't have the cooling capacity to keep the chip under control. 

 

That does not mean Apple is absolved of responsibility. It just means that those others are thrown into the ring of fire (pun not intended) 

Are they suffering from bad power delivery though, as bad temps is something i would expect. Like apple isnt known for good cooling. It is powerdelivery that is the main issue, hot CPUs are commonplace these days anyway

[D13H4RD]

1 minute ago, GoldenLag said:

Are they suffering from bad power delivery though, as bad temps is something i would expect. Like apple isnt known for good cooling. It is powerdelivery that is the main issue, hot CPUs are commonplace these days anyway

XPS15

 

Though I'm not sure if the overheating VRMs due to inadequate cooling count in that case 

[GoldenLag]

15 minutes ago, D13H4RD2L1V3 said:

XPS15

 

Though I'm not sure if the overheating VRMs due to inadequate cooling count in that case 

No doubt these chips run hot when they are turboing. And consume a lot of power while doing so. Seems like good VRMs is crucial for these core i9 chips

[D13H4RD]

1 minute ago, GoldenLag said:

No doubt these chips run hot when they are turboing. And consume a lot of power while doing so. Seems like good VRMs is crucial for these core i9 chips

The i9 requires a lot of power at turbo 

[Stefan Payne]

47 minutes ago, D13H4RD2L1V3 said:

The i9 requires a lot of power at turbo 

So the 45W TDP that Intel mentions is "a bit optimistic"??

[D13H4RD]

4 minutes ago, Stefan Payne said:

So the 45W TDP that Intel mentions is "a bit optimistic"??

Again, that 45W TDP is rated at base speeds on stock settings when the CPU is performing heavy but realistic loads. 

 

It's already been said that going into turbo boost will have it go over. Normally, TDP in a boosted state is not counted on Intel sheets

[Stefan Payne]

3 minutes ago, D13H4RD2L1V3 said:

Again, that 45W TDP is rated at base speeds on stock settings when the CPU is performing heavy but realistic loads. 

 

It's already been said that going into turbo boost will have it go over. Normally, TDP in a boosted state is not counted on Intel sheets

That sounds likey you're trying to defend Intel at all cost, does it?

 

So what's your oppionion on throtteling AM3+ Boards like Asrock 970 PRO3 (IIRC)?

[D13H4RD]

1 minute ago, Stefan Payne said:

That sounds likey you're trying to defend Intel at all cost, does it?

How did you come to that, I have to ask? 

 

Also, Skylake-X? X299 boards with VRM issues? FX-9590?

 

I'm pretty sure AMD and Intel have their share of eggs splat on their faces. Just that AMD is having the upper end now for the desktop side

 

(There's no Ryzen laptop CPU for the high end yet) 

[PeterBocan]

5 minutes ago, Stefan Payne said:

That sounds likey you're trying to defend Intel at all cost, does it?

At this point the discussion is beyond pale. If you can't handle a parachute, don't go paragliding. If you can't handle the load of the top CPU, don't go and buy it. It baffles me that you can not grasp the simple logic. 

[asus killer]

there is also the possibility that people are not using it as intended.

 

[mynameisjuan]

19 minutes ago, PeterBocan said:

At this point the discussion is beyond pale. If you can't handle a parachute, don't go paragliding. If you can't handle the load of the top CPU, don't go and buy it. It baffles me that you can not grasp the simple logic. 

Dont bother explaining to some people. Its honestly not hard to understand. 

[Stefan Payne]

42 minutes ago, D13H4RD2L1V3 said:

How did you come to that, I have to ask? 

Because you admit that the CPU uses more than other 45W TDP CPUs of Intel, you yourself admit that.

Yet you don't want to see the fault from Intel for choosing the 45W TDP for this CPU that looks like a mobile Version of the i7-8700 or even 8700k...

 

42 minutes ago, D13H4RD2L1V3 said:

Also, Skylake-X? X299 boards with VRM issues? FX-9590?

Yes, they also have VRM Issues, even at Default Speeds!

And Skylake-X was a Plattform that came out under the Preassure fo AMD, Kaby-Lake did not.

 

Its the same with this i9 CPU. So why don't you want to admit that the 45W TDP category for this CPU might not have been the wisest choice??

 

42 minutes ago, D13H4RD2L1V3 said:

(There's no Ryzen laptop CPU for the high end yet) 

Yes, because nobody would buy that.

If there was a market for that, it would have been already there. And there already is an ASUS Notebook with an 8 Core Ryzen CPU...

And if more would want that, there would already have been a mobile Version of that, even if AMD had to do a MCM Package with the VEGA Mobile Thing...

 

 

[PeterBocan]

 

[HectorF]

14 hours ago, HammerDrop said:

Not looking to support apple but kinda want to quash this idea that intel gave specs and apple didn't listen. The mobile i7's and i9's are all 45w tdp. You can make the argument that tdp is not a good measure for thermal requirements, but for everyone saying apple didn't build to spec.... well they did. Assuming it was good enough for the i7, it (according to spec sheets and info from intel) should be good enough for the i9. If on the other hand the i7s are running into similar issues, well then yeah peoples points here about apple ignoring the specs would be correct. Most of the reviews I've seen so far indicate that the i7 version is fine. And if you design according to spec, so should the i9. Arguments about, "oh this is the top of the line chip so you should expect it to me amazeballs and require more thermal dissipation" are moot if the spec sheet say 45w tdp. 

 

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

If that is true, then why would the i9 be pushing the VRM any more than the i7s? The fact is that their 3000+ machine is having VRM issues, with no clear solutions other than changing the watts it is drawing. Also in the Reddit, you can see that apple's stock turbo setting is 125W. I wouldn't trust their janky cooling solution to cool anything north of 60W. It is just a short-coming. Apple needs to fix this with more than a 2$ thermal pad. 

[D13H4RD]

Christ dude... 

17 minutes ago, Stefan Payne said:

Because you admit that the CPU uses more than other 45W TDP CPUs of Intel, you yourself admit that.

Yet you don't want to see the fault from Intel for choosing the 45W TDP for this CPU that looks like a mobile Version of the i7-8700 or even 8700k...

Dude, I and 4 others in another thread have already said. Intel's TDP specifications are a calculated average of measurements regarding the CPU's TDP while it is doing intensive loads at base clocks. 

 

If the 8950HK has a 45W TDP at its base 2.9GHz clock, then it is technically valid as that's been defined in the case of Intel. Your TDP will obviously go higher the more you clock. 

 

Also, every CFL-H CPU goes above 45W in turbo. 

 

Sure, we can suggest that Intel puts a boost TDP rating to specify the approximate TDP when the CPU is at max boost. However, note that Turbo Boost was meant to be used in bursts and when power and temp limits allow. Just because you can boost to 4.3GHz doesn't mean you have a 4.3GHz CPU. You still have a 2.9GHz CPU that can overclock itself to 4.3GHz if power and temperatures allow. Furthermore, then clockspeeds vary depending on how many cores are utilized. 

 

17 minutes ago, Stefan Payne said:

Its the same with this i9 CPU. So why don't you want to admit that the 45W TDP category for this CPU might not have been the wisest choice??

Because at base clocks and at heavy but realistic loads, it's specified that the TDP is 45W. If it was 65W, you'd bet that they'd be rating it as a 65W TDP CPU. 

 

17 minutes ago, Stefan Payne said:

And there already is an ASUS Notebook with an 8 Core Ryzen CPU...

The GL702ZC has a desktop chip. I was talking about mobile oriented chips that compete directly against Intel. 

 

Also, because you seem to think I have a hard on Intel (even though I've recommended Ryzen CPUs so many times, I've lost track), I'll leave you with this. 

 

On that laptop, the Ryzen 7 1700's package power reading was 65W during a combined FurMark and Prime95 stress test. And its multicore score on Cinebench R15 was 1400.

[C2HWarrior]

7 hours ago, DrMacintosh said:

A challenger approaches! Another laptop hit with the Core i9 plague

 

Nah this isn't half as bad. For one the CPU performs better than a 4 core while at respectable temps unlike apple and razer also has to cool a much more powerful gpu

[asus killer]

2 hours ago, PeterBocan said:

 

typical mac user, nothing makes sense. Bought a insanely expensive laptop when already had the MBP 2017 and then uses a eGPU and then uses a monitor. Does he even need a laptop?

[Ryujin2003]

90% of MBP users I've met use their devices for email, YouTube, FB, Twitter.. you get the point....

 

Could just be that I haven't met the other side if the coin, but I feel that if the majority of the users don't actually user MBP for production, and don't experience these issues, then will this actually cause an impact to Apple's bottom line?

[PeterBocan]

53 minutes ago, asus killer said:

typical mac user, nothing makes sense. Bought a insanely expensive laptop when already had the MBP 2017 and then uses a eGPU and then uses a monitor. Does he even need a laptop?

Consumerism. They don't need it, but they have to have it. It doesn't matter that it costs the same as trip to Europe, they have to have it, because they feel obsolete.

[EarthWormJM2]

6 hours ago, Stefan Payne said:

So the 45W TDP that Intel mentions is "a bit optimistic"??

Ok, apparently this post, which was specifically about VRM, has gone back to the TDP argument... So I am going to repost the image pulled directly from Intel's website about TDP that @Dylanc1500 posted earlier that appears everyone ignored. It specifically explains that a manufacturer using Intel CPUs has to design their own cooling solution and make sure its adequate enough to run at the correct frequencies. It is not Intel's job to hold everyone's hand when manufacturers have to design their own cooling solutions. I'm not specifically trying to defend Intel here, but it is clearly laid out by Intel and I have read this whole thread and people keep seeming to think Intel did not communicate this to Apple, which is false. And frankly, if a company really needed all its technical information spoon fed to them then they probably wouldn't last very long... Apple is still going strong.