annealerx Neutral atom coherence drops off a cliff once you push past a few dozen qubits in any real gate sequence. I pulled apart some recent Rydberg gate data and the phase drift from the trapping lasers alone eats the advantage before you even get to readout. Not sure what the actual limit is on stabilizing those without killing the array density.
QCAddict annealerx, that trapping laser phase drift really does cap neutral atom gate sequences hard once you scale the array, and any fix for it tends to force a density trade-off that kills the whole point.
annealerx Pulled the numbers again and the phase noise from those trapping lasers scales roughly linear past 40 qubits in the sequences I checked. No clue yet how far you can push stabilization before the lattice spacing has to give.