The Speed-Coherence Dilemma in Quantum Control

You are reading version 1 of this direction, frozen 2026-10-04. The latest version is version 2.

Version 1 of 291 papers119 conceptsfrozen 2026-10-04

In quantum computing, we are caught between two clocks. The first is the computational clock: how fast we can manipulate a qubit to perform a gate or move it across a processor. The second is the decoherence clock: the inevitable countdown until the environment interacts with our system and destroys its quantum information.

Ideally, we would run our computational clock as fast as possible to beat the decoherence clock. However, a recurring theme across decades of research—from early silicon architectures to modern neutral-atom arrays—is that pushing for speed often invites new, aggressive forms of error. This fundamental tension is known as the speed-coherence dilemma.

The Adiabatic Penalty

Many of the most robust quantum operations rely on the adiabatic theorem. This principle suggests that if you change a system's parameters slowly enough, the system will stay in its "safe" ground state throughout the process.

As Victor V. Albert and colleagues explain in Holonomic quantum computing with cat-codes, geometric (holonomic) gates are prized for their noise resilience, but they are fundamentally constrained by this adiabatic requirement. To remain in the computational subspace, the gate must be executed slowly. But as Charles D. Hill noted in Fast nonadiabatic two-qubit gates for the Kane quantum computer, this slowness is a double-edged sword: the longer an operation takes, the more "exposure time" the qubit has to environmental noise and energy relaxation.

The Non-Adiabatic Risk

If slowness is the problem, why not simply turn up the power? The reason lies in leakage and excitations.

When we drive a system quickly, we introduce high-frequency components into our control pulses. These frequencies can inadvertently match the energy gaps of unwanted higher-energy states. In Numerical Simulation of Leakage Effect, Tao Wu and Jianshe Liu demonstrate that "hard" pulses cause the qubit to "leak" out of the computational 0 and 1 states into higher levels, effectively destroying the logic of the operation.

This trade-off isn't just theoretical; it's a hardware engineering bottleneck. R. Trappen and the team at Waterloo showed in Decoherence of a tunable capacitively shunted flux qubit that the high-bandwidth control lines required for fast operations act as "thermal leaks," injecting noise from the outside world directly into the qubit and shortening its relaxation time ($T_1$).

Engineering the Goldilocks Zone

To solve this, researchers are looking for "sweet spots" where speed and coherence are no longer in direct opposition. Miguel J. Carballido and co-authors recently demonstrated a Compromise-Free Scaling of Qubit Speed and Coherence. By operating at a specific "Fast and Coherent Tunable Operating Regime" (FACTOR), they showed it is possible to increase driving strength without a proportional increase in noise sensitivity.

Another approach is to abandon adiabaticity entirely in favor of Shortcuts to Adiabaticity (STA) or non-adiabatic geometric gates. Bao-Jie Liu and colleagues proposed a Plug-and-Play Approach to Non-adiabatic Geometric Quantum Computation that uses complex pulse shaping to achieve the stability of geometric phases at the speed of dynamical gates. The intuition here is like driving a car around a sharp turn: if you go fast, you'll skid (non-adiabatic error), but if you carefully bank the road and steer (pulse shaping), you can maintain high speed without losing traction.

Architectural Specialization

Finally, the field is moving toward heterogeneous architectures to bypass the dilemma at the system level. In Microarchitectures for Heterogeneous Superconducting Quantum Computers, Samuel A. Stein argues that we shouldn't expect a single qubit type to do everything.

Instead, we can use:

  1. Compute Qubits: Highly coupled and fast, but with shorter coherence times, used for rapid logic.
  2. Storage Qubits: Highly isolated and slow to manipulate, but with massive coherence times, used as quantum memory.

By shuttling information between these specialized zones—a process explored by Pau Escofet in Compilation Techniques for Spin Qubits—we can perform fast operations when needed while "parking" the quantum state in a safe haven when it's idle. This architectural shift acknowledges that the speed-coherence dilemma may be a fundamental law of physics, but it is one we can navigate through clever design.

Papers in this version

Showing 60 of 91 papers.

History & citationVersion 2 · 2026-10-04

Current · Permanent link

  1. Version 2 (latest) · 2026-10-04 · 67 papers
  2. Version 1 · 2026-10-04 · 91 papers
Cite this direction

Latest version (a living page):

@misc{lacuna_adiabatic_decoherence_tradeoff_b9d5a34a,
  title = {adiabatic-decoherence tradeoff},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/adiabatic-decoherence-tradeoff/rdr_b9d5a34a8349490d9d27c91aacc96e7f},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_the_speed_coherence_dilemma_in_quantum_c_b9d5a34a_v1,
  title = {The Speed-Coherence Dilemma in Quantum Control},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/the-speed-coherence-dilemma-in-quantum-control/rdr_b9d5a34a8349490d9d27c91aacc96e7f/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}