Why buy a Mac at all if you want to game? We all know that a windows machine is better at it than a Mac laptop with a new ARM chip will be.
Who says that is all you want to do? There are plenty of applications for CUDA, for instance, that don't have to do with gaming.
At the end of the day, people can want to use a $3000+ laptop for more than just one thing. But the difference in what we're seeing here so far - awesome performance in some tasks (namely in some benchmarks and some applications), mediocre performance in other tasks (some other benchmarks, gaming) - is what's a big hangup.
Sigh … there aren’t going to be many games optimized for a GPU that doesn’t exist. And these didn’t in the Mac until recently. Apple compared it to a mobile 3080 because that is what it can be like if a game is actually optimized for it.
A few things: the M1 has been around for nearly a year now, and Metal is used plenty elsewhere. And the optimization excuse only goes so far - the games that are M1 native do perform well, but are still nowhere near 3080 Mobile levels.
And Apple compared it because it is marketing - look at their own chart where it is "near 3080 performance at less power" - what application or benchmark were they using? And how does that translate to real life usage?
To use an analogy, all the PC gamers gnashing their teeth at their expensive, supposedly far more powerful than a console hardware but unable to run some of those popular console ports very well (or at all) because the port did the bare minimum to optimize the game for PC. In this case, it’s even worse because most of the tested games were released years before the shiny new hardware came out. So naturally *can’t be optimized for it*.
Er, what? What PC games - at top end hardware - aren't able to run console ports these days?
Again, "cant be optimized" isn't the case since we've already had the M1 out for a year (which shares the same GPU core architecture), and there are games that *have* had work done to be M1 native. And even so, they're not performing as well at gaming compared to other GPUs versus the difference how they perform in other applications.
Which is fine, since these are different architectures - these Apple GPUs, I'd imagine, would be horrible at ray tracing. But I think when people hear "RTX 3080 mobile levels of performance" they're expecting something other than than RTX 3060 mobile levels of gaming performance.
And to be fair, your argument works the other way too: what if developers optimized games specifically for ONLY an RTX 3080 and used that benchmark to prove the hardware has superior perf/watt to everything else? Would you find that a bit peculiar or even misleading? Would you accept that as a valid benchmark against the M1 Max? We're talking about comparing this against GPUs in the supposedly fractured and unoptimized PC world, no?
That's current state of play. It's taking games and engines designed for an Nvidia and AMD-like GPU and then wondering why it doesn't perform as well on something very different. AMD and Nvidia IMR GPUs are ***much*** more similar to each other than either are to a TBDR GPU like what's found in the M1. 1 year is a piddling amount to get a decent number of actually optimized games. So yeah it's going to take awhile *if ever* because Macs are, for a variety of reasons, not a big priority to optimize for.