[numpad0]

This isn’t necessary for VR because worst tolerated motion to photon latency of 20ms means you can’t be more than 6000km/3700mi from GPU physically assuming zero display and various processing latency, which realistically take 1/3rd each of total latency and eats into budgets accordingly

[gpm]

Also assuming you don't do any post processing tricks locally to accommodate recent movement. I don't think we've explored vr thoroughly enough to know what's actually possible.

Edit: An no reason we can't stick GPUs on an airplane, they aren't heavy.