If this is coming from this article: https://www.apple.com/newsroom/2021/10/introducing-m1-pro-an...

In the picture titled "GPU performance vs. power", that is "relative performance" with lower power consumption, so it doesn't say anything about literal performance.

Then there's this:

> In addition, the GPU delivers performance comparable to a high-end GPU in a compact pro PC laptop while consuming up to 40 percent less power, and performance similar to that of the highest-end GPU in the largest PC laptops while using up to 100 watts less power.[2]

Where that 2 points to:

> ... Discrete PC laptop graphics performance data from testing Lenovo Legion 5 (82JW0012US). High-end discrete PC laptop graphics performance data from testing MSI GE76 Raider (11UH-053). PC compact pro laptop performance data from testing Razer Blade 15 Advanced (RZ09-0409CE53-R3U1). Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro.

  82JW0012US on Amazon has a 3050Ti in it
  11UH-053 has a 3080 in it
  RZ09-0409CE53-R3U1 has a 3080 in it

So there's some data. I've still got some real questions about why they're talking "relative performance" in their graphs, though.

(edit: formatting)

It was written very small in grey text on the bottom right of the slide. I didn't notice it during the presentation but people posted screenshots on Twitter. The "Compact pro PC laptop" is Razer Blade 15 Advanced (rz09-0409ce53) and "High-end PC laptop" is MSI GE76 Raider (11UH-053). You can see if you pause at 22:46[0]. Both of these have RTX 3080s and in Apple's chart their chips are slightly slower but at much lower power. What we don't know if what benchmark they used for this comparison since 375 "relative performance" isn't a defined metric.

[0] https://youtu.be/exM1uajp--A?t=1366

What does the vertical axis mean in those graphs? "Relative performance" seems made up.

Certainly there are tasks they can do to skew that one way or another.

Probably covering their butts in case whatever benchmark they used works differently on ARM vs x86, or if there are weird macOS differences. If there were a bunch of performance patches made to the Mac version of Shadow of the Tomb Raider that made it faster on macOS, that's not because of a faster chip (just an example, I don't know how they benchmark).

There was a link in the slides for the models but I missed it.