One Tunable Knob to Rule Them All: A Unified Two-Qubit Gate Architecture for Neutral Atoms

A single XXZ interaction with adjustable anisotropy, combined with global microwave pulses, can natively produce any two-qubit gate — eliminating the need for multiple hardware-specific interaction schemes.

Imagine you have a piano where every key plays only one note in one octave. You can still play any melody, but you'll need to hit a lot of keys in sequence, transposing constantly. Now imagine a piano where a single pedal smoothly shifts the entire keyboard between octaves. Suddenly, melodies that required elaborate workarounds become one or two keystrokes. That pedal is the tunable XXZ anisotropy in this paper — one continuously adjustable interaction that replaces an entire zoo of special-purpose entangling gates. The committed claim: a single Rydberg-atom implementation of a tunable XXZ spin interaction, combined only with global single-qubit microwave pulses, can natively realize the complete set of two-qubit gates — including B, QFT, iSWAP, √iSWAP, and a continuous CZ-to-iSWAP family. This is not a new entangling mechanism; it is an architectural unification. Instead of engineering separate pulse sequences or hardware configurations for each gate in your instruction set, you tune one parameter (the anisotropy of the XXZ coupling) and apply optimally controlled microwave fields. The ladder question matters here. Current neutral-atom platforms typically implement one or two native gates — most commonly CZ or CPhase — and compile everything else from those primitives. That compilation overhead is real: decomposing an arbitrary two-qubit unitary into CZ plus single-qubit rotations costs additional gate depth, which burns coherence time. The paper's value proposition is that by making the full gate set native, you reduce circuit depth and compilation cost. However, the authors do not report explicit fidelity numbers from hardware experiments or detailed noise simulations — the demonstration is at the level of pulse-optimization theory, not measured gate errors on a real device. Architecturally, this sits in the family of optimal-control approaches to quantum gate design. The XXZ Hamiltonian — H = Jx(SxSx + SySy) + Jz(SzSz) — is a well-known spin model; what is new is the proposal to realize continuously tunable Jz/Jx anisotropy via Rydberg dressing of neutral atoms, then use that tunability as the single primitive for the entire gate set. The optimization of global microwave pulse sequences on top of this interaction is handled via standard quantum optimal control methods. The approach is hardware-specific to neutral-atom (Rydberg) platforms but the algebraic insight — that XXZ with tunable anisotropy plus local unitaries generates all of SU(4) — is platform-agnostic. Integrity is the weak spot. This is a theoretical/computational proposal. The gates are demonstrated via pulse optimization and numerical simulation, not on physical hardware. There is no reported gate infidelity under realistic noise models (spontaneous emission from Rydberg states, laser phase noise, atom motion). The authors do not compare against a specific competing pulse sequence with reported fidelities on a named device. The validation is self-consistent — the optimal control finds pulse sequences that achieve the target unitaries — but the question of whether these sequences survive contact with real decoherence channels is deferred. The milestone to watch is gate fidelity on hardware. The neutral-atom community is currently pushing two-qubit gate fidelities past 99.5% for CZ gates (Lukin group at Harvard, Atom Computing, QuEra). The question this paper must answer next is: does the tunable-anisotropy scheme achieve comparable fidelities for the non-CZ gates (iSWAP, B, QFT) on a real device? If native iSWAP or B gates can hit 99%+ fidelity, the compilation-depth savings become quantitatively meaningful for near-term algorithms. The obvious experiment not run is a noise simulation or hardware demonstration. The honest read: this is likely (a) compute/access limitation — the authors are theorists proposing a scheme for experimentalists to implement — combined with (c) strategic sequencing, where the clean theoretical result is published first to attract experimental collaborators. The absence of noise modeling is notable but not unusual for a theory proposal at this stage.