I don't think this is accurate. [0] and [1] have 3DMark Ice Storm Unlimited Graphics benchmark tested on both the M1 and the 4800U in the Lenovo Yoga Slim-7-14ARE and the M1's GPU stomps the Vega 8 R4000. It outscores even the Ryzen 5000 series iGPU.

I've seen no evidence that the M1's GPU is anything but best in class for integrated graphics.

[0] https://www.notebookcheck.net/Apple-M1-GPU-GPU-Benchmarks-an... [1] https://www.notebookcheck.net/AMD-Radeon-RX-Vega-8-Ryzen-400...

From "doesn't even start to compete" and "GPUs have always been the biggest power draw" I don't think they were comparing to integrated.

But then again Apple GPU were limited to 10W Max and it was a design decision not technology limitation. There is nothing that stops Apple putting in 16 or 32 Core GPU which would bring its GPU performance on par if not exceed market competitors.

> But then again Apple GPU were limited to 10W Max and it was a design decision not technology limitation.

No, it was a technology limitation. The reason I know is because of the M1 Macbook Pro, a device explicitly designed to reap the maximum performance benefits of the M1. It was designed with active cooling and still didn't really manage to score much better than the M1 Macbook Air.

Either way, Apple can't just magically increase the wattage of their chip and make it run faster. They had every opportunity to do that in the M1 Mac Mini and the new iMac, but they didn't. It's a very obvious limitation of the SOC's capabilities, and I honestly can't find any evidence to suggest otherwise.

>a device explicitly designed to reap the maximum performance benefits of the M1.

>Apple can't just magically increase the wattage of their chip and make it run faster.

Because that is not how it works.

The M1 GPU has a maximum TDP of 11W. That is by design. They could push it with higher Clockspeed beyond their optimal level with higher voltage, but that has other testing and reliability implication with cost. It doesn't matter whether you have a large Fan and heat sink sitting on top of or no cooling. You run at 11W Max. That is part of the design. M1 MacBook Air would allow it to run 11W for a fraction of time before heaving TDP headroom to CPU. MacBook Pro allows it to use at its maximum for longer. Since it has a 25W cooling capability.

You want higher GPU performance, throw in more Core. GPU workload are inherently parallel, the only limitation are interconnect and Memory bandwidth. Both of these are are not technical barrier but cost concern.

On a 10W GPU comparison ( And why would you compare GPU that are running at higher watts ) The Apple GPU are doing fairly well in all Metal optimised benchmarks. Compare to AMD Radeon GPU which also has been optimising on Mac platform. ( Although at a larger node )

You're right, but I don't think that the other commenter was speaking to iGPUs. Higher wattage laptops mostly have beefy GPUs sucking up that power, and you'd be hard pressed to find any iGPU that compares (including that of the M1)

Sure, but multiply that by several times (such as the leaks saying that the M1X chip will have 16 or 32 graphics cores compared to the M1's 8) and it'll still be well within the territory of the current 100W spec while handily outperforming pretty much everything else with the same power draw.

Which kind of laptops are you speaking for? M1 GPu performance Blow any integrated gpu and all middle class separate GPUs (mx250, etc).

The mx250 is not a "middle class DGPU" because there is no class below it. It's Nvidia's weakest dedicated graphics card they continue to sell, and it's a pretty terrible point of comparison. A better example would be the GTX 1060, a middle-class dedicated GPU that made it's way into many budget gaming laptops 7 years ago. It is faster than the M1's GPU.

The Mac has a very dedicated audience of video and design professionals who are going to be left empty-handed here, even if they double or triple the amount of GPU cores in the SOC.

Gaming laptops I assume.

M1 GPU beats most other integrated GPUs, but that doesn't magically make the demand for more performance go away, and that demand is likely to increase as there become more and more non-gaming applications for GPUs (machine learning, video editing, etc).

Do you have an actual disagreement? A 14nm Intel chip might waste an extra 20 watts, but when you get into hundred watt territory it's fair to call that a GPU thing.

The Intel astroturf isn’t even artful, it seems either bot or Amazon Turk driven.