The same GPU power that enables 7680x4320@60fps can enable 3840x2160@120-240fps with comparable quality settings.

The same GPU power that enables 3840x2160@240fps with high quality settings might enable 1920x1080@120fps with realtime raytracing instead of matrix-math surface-mapping tricks.

And in any case, "8k" lies at the LOWER end of what a GPU driving a VR HMD really needs to be capable of.

In theory, the eye's "resolution" is commonly stated to be "1 arc second", which works out to 60 pixels per degree FOV. 60-degree FOV is pretty much the absolute lower end of remotely-tolerable tunnel vision. 90-degree FOV is regarded as the minimum anyone really wants to use. 120-degree FOV is approximately the point where the limited FOV is something you're constantly aware of.

For 60-degree FOV, 1 arc-second resolution works out to 3600 pixels. For 90, 120, and 150-degree FOV, it comes out to 5400, 7200, and 9000 pixels. Per eye.

But wait, it gets worse.

The eye's "resolution" is only 1 arc second if you're considering what it can instantaneously resolve. The thing is, your eyes also constantly move around, building up the scene in a manner not unlike synthetic aperture radar. So... to literally eliminate every perceptible hint of a non-antialiased pixel, you have to go even higher. A lot higher. Like, "order of magnitude" higher. Obviously, that isn't happening anytime soon, so we're going to be stuck with things like antialiasing for a long time.

It's also misleading to talk about the eye's "resolution" in its periphery. Yes, indeed, something seen entirely with peripheral vision is comparatively low-resolution and monochromatic... but unless you're either rendering wavefronts onto a contact lens centered on the pupil, you still need to be able to deliver full resolution wherever the eye happens to be looking at that instant. This is the basis behind foveated rendering... combining a display with resolution high enough to give you full resolution wherever the eye happens to be looking, while tracking WHERE the eye is looking to concentrate the GPU's power on that area.

There's a big 'gotcha' with foveation, though... your peripheral vision might be low-res, but it's ALSO extraordinarily sensitive to motion, and your brain subconsciously tracks it to look for "things that aren't quite right". Guess what? Foveated rendering sets off MAJOR subconscious alarm bells. At the very least, it throws off your situational awareness, and you end up kind of like a kid with glasses who has astigmatism & a big strength difference between his left and right eyes trying to catch a baseball... you end up metaphorically getting hit in the face by the baseball, because too many things disrupt the brain's ability to merge and track the object until it's too late.

We have a DEPRESSINGLY long way to go before VR is going to stop feeling "unreal" after more than a few seconds. We need higher resolution displays with faster framerates, ENORMOUSLY faster rendering pipelines to reduce latency, and what I like to refer to as "fog" or "plasma" computing... basically, a cloud-like mini-mainframe sitting a few feet away that's the equivalent of a small rack of i9 motherboards with RTX GPUs. We're not going to get anywhere close to the necessary level of hardware with the equivalent of a top-shelf Android phone (at least, not without tethering it to both power cables AND a refrigerant loop or chilled-water line to keep it from giving you first-degree burns through your clothing).

That's not to say we won't have "usable", and even "fun/enjoyable" VR, before we get to that point... for things like passively watching a 3D movie in a virtual private movie theater, even present hardware is semi-ok. But even then, you'd really want to sit someplace where your head can be held immobile, to avoid breaking the illusion of immersion with lag and "slosh". And after a while, you'll start to feel like you're watching a DVD projected into a screen with a cheap LCoS projector that suffers from major screen-door effect.

Likewise, for spending a few minutes visualizing something like a building's architecture, present-day tech is decent... but an architect will tell you, "VR is a nice place to visit, but I sure as hell wouldn't want to WORK there with today's technology". After a few minutes, the virtual world's cracks become (literally) painfully apparent, and start to really *bother* you.