Cortex A57, Nintendo Switch's CPU

Video games have been around, at this point, for 40-50 years. On that basis alone you'd certainly expect the majority of human video gaming enjoyment to predate the A57.

The Switch has been a commanding platform commercially for its lifespan; given the ease of pushing out multi-platform games with Unity or Unreal Engine, devs are more likely going to target the lower-spec hardware and get a huge install base for free than to limit themselves to more advanced hardware.

Nier: Automata is no. 5 on that list, and while it runs on the Switch, by-all-accounts it's somewhat knee-capped by the hardware. The PS4 original is what's listed, and one wonders if it would've received quite as much adoration if it launched on the Switch and looked or played the way it does on that platform, or even if the devs would have had the resources or freedom to develop it as-it-is in the first place.

Including the top 100, Bloodbourne is at no. 63 and, so far as we know, requires computing power the Switch doesn't provide. Outer Wilds is at no. 96 and you could raise similar questions about it as about Nier; it's made it to Switch (as of last week!), but it took the devs an extra four years to get it there.

It's, frankly, a somewhat quirky list. Among other things, it's heavier than most would be on Japanese visual novels, and it inexplicably has "Lode Runner Online: The Mad Monks' Revenge" at no. 35.

Also, don't underestimate the "regard accumulated over time"/historical import factor, which is a pretty universal element of how critics tend to approach these things. For example, the most recent (2022) Sight and Sound "100 Greatest Films of All Time" has no movies made after 2001 in the top 10; a full half of it predates what most people would consider "modern cinema" (roughly speaking the French and American New Waves in the 60s).

Of course, ultimately your point stands - technology hardly makes the game, and there's a good case that computing hardware has reached a "good enough" point of diminishing returns in the ways that it does. My PS4 is ten years old now and I still haven't seen anything on the next gen systems (themselves around four years old now!) that make me feel like I need to upgrade. I'm also of the opinion that various factors have made games-at-large increasingly less interesting over the past few years, but that's probably just me getting older and sadder.

I suppose the assumption is that you get lots of virtual functions.

While useful for function pointers and jump tables, I suspect they are speaking to polymorphism and vtables/witness tables.

Indirect branches. This is a result of vtable indirections.

You are much more likely to encounter vtables in an object oriented language. Obviously, you can still have the same basic thing in a C program, e.g. SDL RWops, but in C++ for example, it's going to show up all over the place.

FP registers have gotten much larger than the normal types that people store. e.g. 128, 256, 512-bit registers. A normal double floating point (pretty much the largest normally used floating point representation) occupies 64-bits, while a normal int64 occupies, unsurprisingly, the same 64-bits. But we're getting the mega registers specifically because there are a lot of multiple-four-singles at once, and so on, SIMD functions.

> Again I'm pretty certain FP registers are larger owing to the greater precision they have, not specifically because they're designed for vector ops... please somebody explain why my understanding is wrong?

ARM Neon is both 128bit SIMD and the FPU for the system. There's not a separate FPU from the SIMD.

(In some cases the compiler may statically know the class of an object, if it's not allowing for subclassing and a potentially overridden method)

In principle you have calls to whatever method associated with a given object, which are indirect as there is not one statically-known target of the call, but in practise you can optimise a lot out <https://youtu.be/9epgZ-e6DUU?t=2543>.

I can’t see how that would be beneficial.