Stop pontificating.
You have no way of knowing that entanglement violates SR.
For a very, very good reason--you can't set up an experiment that confirms information is communicated when you measure entangled states. You can't postulate a thought experiment either (it would be placed under arrest by the thought police).
But, measurements of entangled photons give random results. How is a random result a communication? Is it like: "I might wire that money to you, or maybe I won't", or "If there are any tapes, I might let you know" . . . ?
By continuing to claim you know something nobody else does, you don't look very scientific, you look more like a crank: ("I alone understand the problem", etc).
Humor me for long enough to do just one more thought experiment.
Let's posit for a moment that relativity is entirely correct and that there is no means of getting a signal across 13 billion light years by any other means than by light propagating from point a to point b.
Now think about the idea that a beam of photons is already there (from point a to b), vastly separated. Think about the idea of a pair of entangled particles that are somehow stationary at either end. Each of the entangled particles "know" from which direction they arrived at point a or point b, respectively.
In this scenario, the particles simply 'stop', and turn about 180 degrees from the DIRECTION they were facing. Has any energy propagated to make this happen? If you observe that either the one at point a or point b has turned, you instantly know what happened to the other entangled twin. How long does it take for a different direction to be chosen with respect to the one from which a photon arrived, anywhere in a universe permeated by an inertialess quantum field that is not spinning? "Not spinning" or changing direction is a kind of inertia also.
If the quantum field in which the mechanism that produced the particles is sensitive to direction, AND if the entangled photons have traveled as they do in a straight line from one end of the universe to the other and are still entangled, then entanglement or 'disentanglement' is nothing more exotic than turning around (changing direction).
This somehow makes perfect sense to me. I don't understand why, but it does. Reducing 'real' simultanaeity to requiring only one observer and one event makes sense as well.
And it does look like something that, in a limited sense, could be made use of to build an extremely fast computational engine, which we know is your focus. I'm sorry that I can't provide any further clues as to how large a qubit might be, but this looks like something more analogous to conventional computing, but at much "higher" / "lower" speeds of computation locally.
It seems to function more like an accumulator (the computational 'action' register) than a deep memory device, so no doubt the first ones will need to interface to more conventional deep memory, if only to keep pairs of entangled qubits straight. The conventional memories will need to keep track of at least three states; entangled 'up', entangled 'down', and 'not entangled'. Worse, this memory transfer interface can't possibly happen without losing information, unless it is 'cloned' somehow.
In the strictest sense, this is the domain of engineering, not science. It has an element of inspiration.
I never had any tapes; just a useless idea about a change of direction being the basis of quantum entanglement.