Well, no - they immediately followed the discrete laptop graphic comparison with a desktop graphic comparison, highlighting how much more power they draw for "the same" performance.
> Well, no - they immediately followed the discrete laptop graphic comparison with a desktop graphic comparison
pretty sure the comparison was with "the most powerful PC laptop we could find", which makes sense because they then talked about how much it was throttled when running only on battery while the new M1 is not.
That wasn't for desktop graphics, it was for a top-end laptop graphics SKU (I think RTX 3080 Mobile on a ~135W-150W TDP configuration?). Otherwise the graph would extend all the way to 360W for a RTX 3090.
And I think based on these numbers that a desktop 3090 would have well over double the performance of the M1 Max. It's apples to oranges, but lets not go crazy just yet.
Now, I am extremely excited to see what they will come up with for the Mac Pro with a desktop thermal budget. That might just blow everything else by any manufacturer completely out of the water.
The chart only shows a line up to about 105W, so it's not clear what they're trying to represent there. (Not that there's any question this seems to be way more efficient!)
