M2 smokes 13th gen Intel Raptor Lake P mobile CPU in single core benchmarks?

[Gudi]

You are aware that AS devices run at similar temps to many x86 based laptops?

Are you aware that AS laptops have become as performant as Intel desktops at the same power consumption? This peak performance bundled with 4 efficiency cores makes it far less likely that you will ever push the machine to the temperatures measured by testers after running Cinebench back to back for 30 minutes. As a user you experience much longer battery life, which means doing the same tasks the M1 will turn way less electricity into heat. Every percentage of battery life left is heat not yet produced.

On the other hand AS desktops have become as power efficient as Intel laptops, which means they can now be cooled by tiny little laptop fans. iMacs used to double as electric heaters in the winter, but not anymore. In the past you could fry an egg or grill a steak on an x86 cpu. This won't work on an M1 iMac sipping 15W. The 24" iMac is just a little bit faster than an 8-core i9 iMac, but it is so much thinner, cooler and quieter.

 

[falainber]

This is far from being true and neither is the full lineup covered in your link. The i7-1370P typically has its TDP set between 25 and 30W in contemporary devices. For 15-25W TDP devices Intel has the U series like the i7-1255U - which is below the P-series.
Geekbench Score for the Dell Inspiron 14 7420 - which has a i7-1255U set to 23W PL1 - is 1640 - which is below M1 while still having a higher PL1.
And below this we have the 9W series like the i7-1250U. This is finally the Intel CPU you can put into a fan-less device and compare to the Macbook Air.

​

That's just one way to compare the processors. OP did not actually specify that power consumption was important for him/her. Another way to compare is to see what's available at given performance or price or some other set of constraints. While I don't have any numbers, anecdotally, the prevalent (by a huge margin) laptop usage mode nowadays is being connected to the power outlet. With that in mind, all these arguments about power efficiency seem pretty pointless for most use cases.

 

[Lihp8270]

Are you aware that AS laptops have become as performant as Intel desktops at the same power consumption? This peak performance bundled with 4 efficiency cores makes it far less likely that you will ever push the machine to the temperatures measured by testers after running Cinebench back to back for 30 minutes. As a user you experience much longer battery life, which means doing the same tasks the M1 will turn way less electricity into heat. Every percentage of battery life left is heat not yet produced.

On the other hand AS desktops have become as power efficient as Intel laptops, which means they can now be cooled by tiny little laptop fans. iMacs used to double as electric heaters in the winter, but not anymore. In the past you could fry an egg or grill a steak on an x86 cpu. This won't work on an M1 iMac sipping 15W. The 24" iMac is just a little bit faster than an 8-core i9 iMac, but it is so much thinner, cooler and quieter.

View attachment 2122714

You wasn’t talking about performance. You was talking about the heat causing issues.

AS still produces the same heat.

 

[Lihp8270]

Don’t confuse temperature and heat.

Do elaborate?

 

[leman]

Do elaborate?

Heat is not the same thing as temperature. Let's take the CPU as an example. M1 consumes 4x less power to produce performance comparable to an Intel CPU. This also means 4x less heat. That's why you can have a passively cooled M1/M2 outperforming actively cooled Intel and AMD CPUs. The chip temperature ends up being the same, but that's simply because Apple can afford to run a smaller cooling solution and/or cool more conservatively.

And in the situations where the power consumption/heat output are indeed the same (sustained operation for example), you get significantly more performance.

 

[Gudi]

You wasn’t talking about performance. You was talking about the heat causing issues.

AS still produces the same heat.

No it doesn't. It can produce the same heat (as an Intel laptop chip), if you throw more demanding tasks at it. But it will run through them (as fast as an Intel desktop chip) without throttling. Under the same workload (typical for laptops) it will not get nearly as hot and leave the energy in the battery instead of transforming it into heat.

 

[Lihp8270]

Heat is not the same thing as temperature. Let's take the CPU as an example. M1 consumes 4x less power to produce performance comparable to an Intel CPU. This also means 4x less heat. That's why you can have a passively cooled M1/M2 outperforming actively cooled Intel and AMD CPUs. The chip temperature ends up being the same, but that's simply because Apple can afford to run a smaller cooling solution and/or cool more conservatively.

And in the situations where the power consumption/heat output are indeed the same (sustained operation for example), you get significantly more performance.

I’m not saying anything about power consumption, or performance though.

 

[leman]

I’m not saying anything about power consumption, or performance though.

Then I am not sure what you are saying. Power consumption and heat are pretty much the same thing.

 

[Lihp8270]

Then I am not sure what you are saying. Power consumption and heat are pretty much the same thing.

A 35W chip can run at the same temperatures as a 200W chip. If the cooking systems have been designed in such a way that this is the case.

I have never said anything bad about AS, there’s no arguing about its performance or efficiency.

The posts I was replying to were those which claimed x86 chip temperatures caused failures. It was nothing about performance, purely chip temperatures.

I simply commented that it’s not true, and that under equally heavy loads AS can run as hot as x86 as that’s how the cooling system is designed. There’s nothing wrong with that and doesn’t cause any lifespan issues.

I was simply pointing out the incorrect assertion that x86 cpu temps alone caused failures.

 

[quarkysg]

A 35W chip can run at the same temperatures as a 200W chip. If the cooking systems have been designed in such a way that this is the case.

You seem to equate heat with temperature. One is measuring energy while the other is a transient measurement of that energy. Related but not the same.

A 35W heating element can boil the same amount of water as can a 200W heating element but the 35W one will take a lot longer.

 

[leman]

A 35W chip can run at the same temperatures as a 200W chip. If the cooking systems have been designed in such a way that this is the case.

I have never said anything bad about AS, there’s no arguing about its performance or efficiency.

The posts I was replying to were those which claimed x86 chip temperatures caused failures. It was nothing about performance, purely chip temperatures.

I simply commented that it’s not true, and that under equally heavy loads AS can run as hot as x86 as that’s how the cooling system is designed. There’s nothing wrong with that and doesn’t cause any lifespan issues.

I was simply pointing out the incorrect assertion that x86 cpu temps alone caused failures.

Ah, sorry if I misread the context. I thought the conversation was about performance and thermal constraints (and also how windows gaming laptops need to accommodate more heat in general). Didn’t want to be annoying.