Apple’s huge cash reserves are by far the least interesting advantage it can bring to bear in this area. There are two and a half more interesting advantages: t

Apple’s huge cash reserves are by far the least interesting advantage it can bring to bear in this area. There are two and a half more interesting advantages: their superstar translation layer developer team[1], their superstar silicon team[2], and the ability for these teams to work together[2.5]. 1: Rosetta 2 was low-key a modern miracle: Apple made an x86-to-ARM translation layer that Just Works for 99% of software. 2: Apple Silicon was another modern miracle, although much more high-key: Apple stepped into chip design for laptops/desktops and produced a thoroughly competitive set of chips (that are even industry-leaders in some, though not all, important metrics). 2.5: One of the reasons Rosetta 2 worked so well was that M-chips were designed with built-in tricks to support it (https://twitter.com/ErrataRob/status/1331736203402547201 https://twitter.com/ErrataRob/status/1331736203402547201). Using these advantages to support gaming is the way to go, not trying to pay game developers to use Metal. Take your Rosetta dev team and task them with making a DirectX-to-Metal translation layer (you even have a ready-made name for it: the Ezana Stone is another trilingual stone like the Rosetta Stone, this time written in Classical Ethiopic, Sabaic, and Ancient Greek). Take your Apple Silicon team and task them with making a gaming-focused chip (this might be as simple as making an M-chip that has more like a 3:1ish ratio of GPU to CPU cores rather than the current 1:1ish ratio). Have these two teams work together to put features on the gaming chip that support the translation layer (explicit hardware support for a few of the trickiest DirectX calls or something). Maybe you can even punt all that fancy neural-network-based DLSS upscaling to the Neural Engine, who knows? The M1 Ultra was essentially two M1 Maxes taped together and much was made of it being “beating” Nvidia’s 3090 (only if you starve the 3090 of power). But tape two M1 Ultras together, and it looks like you probably beat a 3090 in a fair fight too. If Apple wanted to go that way, they could ship a hypothetical system that packs two GPU-core-heavy M3 Ultras taped together, executes a translation layer that runs DirectX/Vulkan alongside OpenGL/Metal, and shoves the pixels onto studio-grade 4K/5K/6K Retina displays. That would be the most powerful gaming PC in absolute as well as relative terms. Apple would go from “by far the worst gaming platform” to “by far the best gaming platform” overnight. I doubt Apple has any interest in this direction, but if they did, they could do so much better than throwing cash at game developers.