Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

leman

macrumors Core
Oct 14, 2008
19,521
19,678
Emphasis mine above. Sure, if someone copies the MBP design it will have the same characteristics - the physics is the same in both cases. The discussion is whether the form factor makes sense (other than aesthetically).

Why wouldn't it? It's a CPU that is supposed to run at 45 watts continuous power dissipation. The chassis is perfectly able to guarantee that. Sure, the same CPU could run faster in a larger chassis with more capable cooling solution. But than we are talking about an entirely different class of computers. You can get a portable machine that will run the same hardware 10-20% faster, at a substantial cost to size, weight, ergonomy and battery life.

I still think Apple's goal is largely aesthetic: thinness + quietness.

I think that Apple's goal is an overall balance of feature in an ergonomic package. Just look how the industry has changed over the last couple of years. Other manufacturers "solve" the same problem by downgrading the longterm CPU TDP. Just a short time back 15W was reserved for ULV CPUs and people were ranting about those "underpowered designs". Now it's the standard TDP for a Windows laptop.
 
  • Like
Reactions: Nightfury326

Argon_

macrumors 6502
Nov 18, 2020
425
256
I am a bit confused, since your screenshot definitively shows a temperature-throttled machine. The cooling system is carefully engineered to keep the chip at it's maximal safe operating temperature when the chip is drawing 30 watts in sustained operation. It can run in this mode for basically forever, showing that the cooling system is adequate. This is exactly how any 15" and 16" Intel Mac operates as well. I though you were arguing agains this type of power management?

A power-throttled machine looks different: you would most likely see lower CPU temperatures while the machine bounces of the PL-levels instead of the temperature.

By the way, what your screenshot doesn't show is the initial frequency spike before the system has settled. Try setting the Power Gadget sampling rate a bit lower, and start the test again. You will see the chip immediately draw more power (probably closer to 50 watts) and run on higher frequencies, bouncing off 100C after a couple of seconds and then settling at around 95C/28W. That's how Apple thermal management works on Intel Macs.
I too am confused.

My screenshot clearly shows a CPU following a programmed power curve, as opposed to reaching a hundred degrees and turning the tap back. The CPU did not reach its T junction during the benchmark. Note that the required and actual frequencies match. They would not if the machine were throttling.

Look at the screenshot from my earlier post. The one sourced from MaxTech. It clearly shows two different situations. One involving a programmed power curve, the other involving throttling.

My argument is in favor of programmed power caps, as opposed to the roast and throttle approach. At least for machines with PRO written on them.
 

Ethosik

Contributor
Oct 21, 2009
8,142
7,120
And yet Macs have consistently cooler chassis than most other laptops out there. And they are generally quieter. Look up notebookcheck reviews, they do measurements. It's not uncommon for a Windows laptop with a 45W class CPU to reach underside temperatures of over 50C — because they push the air out through the bottom chassis. Mac bottom chassis is on average around 10C cooler.
Yeah I agree with this. I have an HP laptop for work and it get absolutely hot and the fans also get loud. But people like to just think Apple is the evil ones. My mac laptop actually is more cooler than my HP work laptop.
 
  • Like
Reactions: Nightfury326

Ethosik

Contributor
Oct 21, 2009
8,142
7,120
Sure we do. It's making money (ie keeping their margins on products at certain levels) and others things that can only be talked about in the PRSI forum.
Can we please stop this discussion about Apple trying to make money in almost every topic (not directed at you, just using your comment as a reference point).

Look, Apple is a Company. ALL companies top priorities are to make money. A company solely exists to make money. But there are companies out there that have good secondary/tertiary goals - like hospitals. Yes Apple wants to make money. But I believe throughout the history of Apple that they want to make great products and also protect our privacy.


And what people fail to realize when they compare an Apple computer to a Dell/HP/Lenovo is EVERYTHING. Not just CPU vs CPU or RAM vs RAM. But what about resolution? Build quality? How well it performs unplugged? macOS vs Windows? iWork and Garageband included?

Some of those are subjective yes, but almost everything is. Does everyone on the planet need 16GB instead of 8GB? No.

When properly compared, the pricing is actually VERY close to the windows equivalent systems. It can be $100 or $200 more, but that is the decision you need to make as a purchaser.
 
  • Like
Reactions: Nightfury326

Andropov

macrumors 6502a
May 3, 2012
746
990
Spain
I too am confused.

My screenshot clearly shows a CPU following a programmed power curve, as opposed to reaching a hundred degrees and turning the tap back. The CPU did not reach its T junction during the benchmark. Note that the required and actual frequencies match. They would not if the machine were throttling.

Look at the screenshot from my earlier post. The one sourced from MaxTech. It clearly shows two different situations. One involving a programmed power curve, the other involving throttling.

My argument is in favor of programmed power caps, as opposed to the roast and throttle approach. At least for machines with PRO written on them.
But... why? Your comment seems to suggest that Apple just slaps Intel's CPUs in the chassis and lets them auto-throttle themselves. Which isn't true. Apple knows the thermal limitations of the chassis they design, but they prefer to run them closer to the max safe CPU temperature because there's a thermodynamic efficiency gained in doing so and nothing is really lost.

Let's make a thought experiment. Imagine you wanted to use air cooling to keep two identical CPUs (A and B) at 100ºC and 60ºC each, both consuming the same amount of power (i.e. 45W). You would see that keeping the second CPU at 60ºC requires more airflow (for example, 0.5 m³/s instead of 0.3 m³/s) since the heat is transferred faster (from the heatsink to the cooling air) proportional to the difference in their temperatures. The higher the difference, the less airflow you require. Since more airflow means more noise, Apple simply chooses to run the CPU at 100ºC and have less noise.

Typically, Apple uses the approach of processor A and PC makers the approach of processor B. You could make the cooling system of A bigger, so it runs at 60ºC without needing to increase the airflow (and fan noise), keeping the airflow at the 0.3 m³/s in the example. But what would you achieve? You could argue that it allows for higher sustained performance if you let the temperature go beyong 60ºC, but then the sky is the limit, you would end up designing a desktop PC if you don't take other factors ito account.

The way Intel CPUs are designed now, they will take as much power as you let them, while getting diminishing returns for the power used (since power consumption scales quadratically and performance only scales linearly). But once you settle in a certain chassis, running it at 100ºC allows for maximum performance.

The only exception may be at low frequencies / low load, where the fan is almost inaudible anyway and upping the frequency slightly to avoid the laptop being too hot to the touch is a better tradeoff.
 

leman

macrumors Core
Oct 14, 2008
19,521
19,678
Look at the screenshot from my earlier post. The one sourced from MaxTech. It clearly shows two different situations. One involving a programmed power curve, the other involving throttling.

Can you provide more info about the source? Like the original article or video? It does seem that there was some frequency fluctuation in the MaxTech machine, which could be down to one of the following:

- a faulty or buggy unit
- a flawed test methodology
- different OS version with a different power management algorithm

No MBP I have tested (and I've tested a bunch) showed this kind of fluctuations under controlled circumstances by the way, so I'd speculate the problem here was on MaxTech end.
 

pshufd

macrumors G4
Oct 24, 2013
10,149
14,574
New Hampshire
Read the MacBook Pro forum for the 2016-2019 Intel MacBook Pros. The two biggest complaints are the keyboards and heat. Sure, you can hit those high frequencies for performance but you can't hold it for long. Intel really hasn't done anything for four years and Rocket Lake is more of the same - you can get more performance at really high power drain. Instead, look at what AMD has been doing - AMD has been showing up Intel quite nicely for a few years. Intel is still selling CPU parts but some of that is due to AMD parts being sold out. Intel has been a major disappointment for over a decade. In the meantime, Apple has been doing great things with their A series chips.

Intel used to be on Tick-Tock and that's what Apple was counting on. New process every two years, new microarchitecture every other year. It worked until it didn't. Why do you think Microsoft tried Windows on ARM a few years ago? They have the same issue.
 
  • Like
Reactions: Nightfury326

steve1960

macrumors 6502
Sep 23, 2014
293
300
Singapore
Apple processors work better under certain circumstances than Intel processors. That was always going to be the case given Apple designed the processor for their own optimal use case. That is what companies do when they design their own chip. The Intel chip is designed to be an all encompassing chip for all circumstances.

Liken this scenario to a car. The standard chip in the cars computer is Intel. It is able to function globally across all fuel types and grades. The custom chip in Apple's car is optimized for the best grade of fuel and pulls out the very best perforate and optimists everything.

However, there are still compromises because no solution is everything to everyone, well at least not yet!
 

pshufd

macrumors G4
Oct 24, 2013
10,149
14,574
New Hampshire
Apple processors work better under certain circumstances than Intel processors. That was always going to be the case given Apple designed the processor for their own optimal use case. That is what companies do when they design their own chip. The Intel chip is designed to be an all encompassing chip for all circumstances.

Liken this scenario to a car. The standard chip in the cars computer is Intel. It is able to function globally across all fuel types and grades. The custom chip in Apple's car is optimized for the best grade of fuel and pulls out the very best perforate and optimists everything.

However, there are still compromises because no solution is everything to everyone, well at least not yet!

That's part of it. But a big part of it is VLIW and that's impossible to fix with x86. Intel and AMD are just at a serious architectural disadvantage. Kudos to Apple for taking advantage of it in a high-power process.
 

steve1960

macrumors 6502
Sep 23, 2014
293
300
Singapore
And here it is. Semiconducter architecture. The black art! As you say Intel and AMD have missed the forward thinking because they have been supplying generic products for too long. Custom silicon has been around for a long time and has a proven business model if executed well (if not it will be expensive and probably bankrupt a company).

Go safe or go for it? Apple decided to go for it as many others have.
 

pshufd

macrumors G4
Oct 24, 2013
10,149
14,574
New Hampshire
And here it is. Semiconducter architecture. The black art! As you say Intel and AMD have missed the forward thinking because they have been supplying generic products for too long. Custom silicon has been around for a long time and has a proven business model if executed well (if not it will be expensive and probably bankrupt a company).

Go safe or go for it? Apple decided to go for it as many others have.

I'm wearing a Garmin Fenix watch. It is a wildly popular watch for Hikers, Runners, Triatheletes, etc. They cost considerably more than the AW. I'd say that Garmin has been quite successful with it. Apple has been successful at it for a long time in iPhones and iPads. They have a considerable track record at it. You could say that the M1 is evolution over revolution.

It has nothing to do with generic products for too long. It's just the architecture.

I don't see why you think that it's a black art. Anyone with a EE or CS degree would understand this stuff.
 

steve1960

macrumors 6502
Sep 23, 2014
293
300
Singapore
Oh and it not just about the silicon, seems Apple got it right. 15 years ago I worked for a company at the leading edge of Bluetooth why did we lose the biggest mobile phone account at the time? the silicon was great, the software was sadly lacking. it is all about optimization not only the silicon.
 
  • Like
Reactions: pshufd

steve1960

macrumors 6502
Sep 23, 2014
293
300
Singapore
I'm wearing a Garmin Fenix watch. It is a wildly popular watch for Hikers, Runners, Triatheletes, etc. They cost considerably more than the AW. I'd say that Garmin has been quite successful with it. Apple has been successful at it for a long time in iPhones and iPads. They have a considerable track record at it. You could say that the M1 is evolution over revolution.

It has nothing to do with generic products for too long. It's just the architecture.

I don't see why you think that it's a black art. Anyone with a EE or CS degree would understand this stuff.
Have you ever taken a deep dive into defect density, mold compound issues and how they affect chip cracking under load, how mold compounds can affect chip performance and yield. Temperatures for curing? I could go on and on. This sh@t is not easy.
 

pshufd

macrumors G4
Oct 24, 2013
10,149
14,574
New Hampshire
Have you ever taken a deep dive into defect density, mold compound issues and how they affect chip cracking under load, how mold compounds can affect chip performance and yield. Temperatures for curing? I could go on and on. This sh@t is not easy.

I have no doubt that it is.

But those things are irrelevant on the main reasons for the performance and performance per watt on AS.
 

steve1960

macrumors 6502
Sep 23, 2014
293
300
Singapore
Relevant. Custom silicon, with an optimized software suite, will always outperform generic silicon from Intel and AMD of elected properly. Comes with a huge R&D overhead but Apple can afford it.
 

pshufd

macrumors G4
Oct 24, 2013
10,149
14,574
New Hampshire
I think you ae probably right but honestly I have never used Apple's cloud I prefer local backups.

Backup is only one aspect of iCloud. I only use local backups as well but I depend on iCloud Notes, Reminders, Calendar, File Sharing, email, and Numbers. Some depend on Photos or the other pieces of it.
 

pshufd

macrumors G4
Oct 24, 2013
10,149
14,574
New Hampshire
Relevant. Custom silicon, with an optimized software suite, will always outperform generic silicon from Intel and AMD of elected properly. Comes with a huge R&D overhead but Apple can afford it.

Sure.

But a really big piece is VLIW.

Do you understand why VLIW is such a big issue in performance?
 

steve1960

macrumors 6502
Sep 23, 2014
293
300
Singapore
Backup is only one aspect of iCloud. I only use local backups as well but I depend on iCloud Notes, Reminders, Calendar, File Sharing, email, and Numbers. Some depend on Photos or the other pieces of it.
Yep I get it and that's good for you. I prefer to have everything under my control. Probably just my OCD kicking in!
 

pshufd

macrumors G4
Oct 24, 2013
10,149
14,574
New Hampshire
Yep I get it and that's good for you. I prefer to have everything under my control. Probably just my OCD kicking in!

The thing is a lot of novice users benefit from these applications having their stuff on the cloud that they can use on their Macs, iPhones, iPads and even Windows systems. There's a big level of convenience about accessing your data or apps on whatever device you have to be using at the time.

It's nice that consumers don't have to install programs - they come with macOS, iOS, iPadOS, and have the flexibility of the cloud.

I personally only put stuff on iCloud where I'm not that concerned about privacy. Stuff where I'm concerned about privacy only goes on local storage. But most people aren't aware of the distinction. I believe that iCloud is a part of the Apple easy-to-use story.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.