Improved Fermionic Scattering for the NISQ Era

This paper proposes a shallow-circuit fermionic scattering state preparation method that approximates wave packets via spatial localization to reduce circuit depth by nearly half while preserving anti-commutation relations, demonstrating its efficacy through MPS simulations and successful implementation on IonQ's Forte 1 NISQ hardware.

The Gist

Imagine you are trying to simulate a high-speed collision between two tiny, invisible particles on a computer. In the world of quantum physics, this is called scattering. It's the fundamental way we understand how particles interact, bounce off each other, or create new things.

However, we are currently living in the "NISQ Era" (Noisy Intermediate-Scale Quantum). Think of current quantum computers as high-performance race cars with very leaky tires. They are incredibly powerful, but they lose energy (decoherence) and make mistakes (noise) very quickly. Because of this, you can only drive them for a very short distance (a "shallow circuit") before the tires go flat and the simulation crashes.

The Problem:
To simulate a particle collision, you first need to prepare the "starting state"—creating two wave packets (like two surfer waves) that are ready to crash into each other.

• The Old Way: The standard method to create these waves was like trying to build a house by laying every single brick one by one, in a single long line. It was accurate, but it took too many steps (circuit depth). On our leaky-tire quantum cars, this process would take so long that the car would break down before the house was even built.
• The Goal: We need a way to build these starting waves faster, using fewer steps, without losing accuracy.

The Solution: "The Localized Shortcut"
The author, Michael Hite, proposes a clever new method that combines two existing ideas to speed things up by nearly 50%. Here is how it works, using some everyday analogies:

1. The "Split-Team" Strategy (Localization)

Imagine you are organizing a massive dance party for 100 people (the particles).

• The Old Method: You had to line everyone up in one giant circle and teach them the dance moves one by one, from person 1 to person 100. This took forever.
• The New Method: You realize that the two waves of dancers only need to interact with their immediate neighbors. So, you split the room in half. You teach the left group of dancers their moves while simultaneously teaching the right group.
• The Result: Because you are doing two things at once (parallel processing) instead of one long line, you finish the preparation in half the time. The paper calls this "localizing" the wave packets. It's like realizing you don't need to talk to the whole room to get your message across; you just need to talk to the people sitting next to you.

2. The "Magic Elevator" (Ladder Operator Block Encoding)

In quantum physics, creating a particle is like trying to push a heavy box up a steep hill. Sometimes, the math says you can't do it directly because it breaks the rules of the universe (non-unitary operations).

• The Old Problem: Trying to push the box directly often results in it sliding back down or breaking the hill.
• The New Solution: The author uses a "Magic Elevator" (Block Encoding). Imagine you have a special elevator that can take the box up to a higher floor (a larger, safe space), move it where you need it, and then bring it back down.
• The Catch: You have to check the elevator's control panel (measure an "ancilla" qubit) to make sure the box didn't fall out. If the panel says "All Good," you keep the result. If not, you try again. This ensures the physics rules are never broken, even though we are taking a shortcut.

3. The "Test Drive"

The author didn't just theorize this; they actually tested it.

• The Simulation: They used a super-accurate classical computer (like a flight simulator) to prove that their "shortcut" method produces waves that look almost exactly like the real thing, especially when the particles aren't crashing too hard (weakly interacting).
• The Real Hardware: They then ran the experiment on a real quantum computer (IonQ's Forte 1). Even with the "leaky tires" of today's technology, the machine successfully created the scattering state with very low error rates (less than 17% error, and only 7.5% if you ignore the quietest spots).

Why This Matters

This paper is like finding a new, faster route through traffic for quantum computers.

• Before: We could only simulate very simple, small crashes because the journey took too long.
• Now: By cutting the preparation time in half, we can simulate more complex scenarios and get closer to the "critical" moments where particles interact in interesting ways.

In a nutshell:
The author figured out how to build the "starting line" for a quantum particle race much faster by splitting the work in half and using a clever "elevator" trick to follow the rules of physics. This allows us to run more complex simulations on today's imperfect quantum computers, bringing us one step closer to solving real-world problems in materials science and particle physics.

Technical Summary

1. Problem Statement

In the Noisy Intermediate-Scale Quantum (NISQ) era, digital quantum computers are severely constrained by system noise and qubit decoherence. These limitations restrict simulations to shallow circuits, typically on the order of 1,000 gate layers.

• The Bottleneck: Simulating fermionic scattering requires preparing complex scattering states (wave packets). Standard state decomposition methods (e.g., Möttönen or Quantum Shannon) result in circuit depths that scale poorly with system size, making them infeasible for current hardware.
• The Specific Challenge: Fermionic wave packets are sums of highly non-local, non-unitary creation operators. Preparing these states efficiently while preserving fermionic anti-commutation relations and minimizing circuit depth is a critical hurdle for real-time scattering simulations.

2. Methodology

The author proposes a modified scattering state preparation method that combines two existing techniques with novel approximations tailored for weakly interacting systems.

A. Localized Wave Packet Approximation

• Concept: Instead of preparing wave packets that span the entire lattice (which requires $O(N^2)$ terms), the method truncates the wave packet sums to act only on specific spatial sub-regions (e.g., left and right halves of the lattice).
• Mechanism: A Gaussian wave packet centered at $x_A$ is truncated to act only on the relevant half of the space.
• Benefit: This allows the preparation of two wave packets to occur in parallel rather than in series.
  • Original Circuit Depth: $8(N-1)$ CNOT gates (series).
  • Proposed Circuit Depth: $\approx 4(N/2 - 1)$ CNOT gates (parallel).
  • Result: A reduction in circuit depth by nearly 50%.
• Validity: This approximation is rigorously justified in the perturbative weakly interacting regime ($g \ll h/2$), where truncation effects are suppressed.

B. Ladder Operator Block Encoding (LOBE)

• Problem: The creation operator $\hat{c}^\dagger$ (or lowering operator $\sigma^-$ in spin language) is non-unitary, which cannot be directly implemented on a quantum computer.
• Solution: The paper employs the Ladder Operator Block Encoding method (Simon et al.).
  • A non-unitary operator is embedded into a larger unitary $U$ using an external control qubit and an ancilla qubit.
  • The system is post-selected on the ancilla being in the $|0\rangle$ state to ensure the correct non-unitary operation is applied.
• Implementation: This allows for the preparation of wave packets in critical theories (where particle number is not strictly conserved in the same way) while maintaining fermionic anti-commutation relations.

C. Simulation Framework

• Model: A modified 1+1D Transverse Field Ising model with a four-fermion interaction term (representing the Thirring model).
• Ground State: Prepared using the Variational Quantum Eigensolver (VQE) with an efficient-SU(2) ansatz.
• Classical Verification: Matrix Product State (MPS) simulations using the ITensor library, Time-Evolving Block Decimation (TEBD), and Density Matrix Renormalization Group (DMRG) were used to benchmark the approximations.
• Hardware Execution: Implemented on IonQ's Forte 1 trapped-ion quantum computer.

3. Key Contributions

1. Circuit Depth Reduction: Introduced a spatial localization strategy that reduces the CNOT depth of fermionic scattering state preparation by nearly half, making it viable for NISQ devices.
2. Hybrid State Preparation: Successfully integrated Givens rotations (for state rotation) with Ladder Operator Block Encoding (for non-unitary creation), enabling the preparation of scattering states in critical theories.
3. Hardware Demonstration: Provided one of the first experimental demonstrations of fermionic scattering state preparation on modern hardware (IonQ Forte 1), moving beyond purely theoretical or small-scale classical simulations.
4. Error Analysis: Quantified the trade-off between approximation (truncation) and accuracy, demonstrating that errors remain low (<10-17%) in weakly interacting regimes.

4. Results

Classical Simulation (MPS)

• Accuracy: In the weakly interacting regime ($g=0.01$), the "truncated unitary" wave packets showed an average relative error of < 2% in occupation numbers and < 3% in entanglement entropy compared to exact wave packets.
• Criticality: As the system approaches criticality (stronger interactions), the perturbative window shrinks, and errors increase (up to ~15% for $g=0.05$), but the method remains a reasonable approximation.
• Comparison: The truncated unitary packets (using LOBE) performed comparably to truncated non-unitary packets, with minor deviations attributed to trace leakage.

Hardware Execution (IonQ Forte 1)

• Setup: An 8-site scattering state was prepared on top of a VQE ground state ($J=0.4, h=1, g=0.1$).
• Success Rate: Out of 5,000 shots, 4,536 successfully measured the ancilla in the $|0\rangle$ state (post-selection).
• Fidelity:
  • The relative percent error between the IonQ-prepared state and the exact classical state was 17.2%.
  • When low-occupation sites (noise-prone areas) were neglected, the error dropped to 7.53%.
• Conclusion: The results demonstrate that accurate initial scattering states can be generated on current hardware with limited error correction and sample sizes.

5. Significance

• NISQ Viability: This work provides a concrete algorithmic pathway for performing real-time scattering simulations on current and near-future quantum hardware, which was previously considered too deep for NISQ devices.
• Scalability: By reducing circuit depth and enabling parallel execution, the method scales more favorably with system size, addressing a primary limitation of previous fermionic simulation proposals.
• Foundation for Future Work: The success of this method on IonQ hardware paves the way for simulating more complex theories (e.g., Gross-Neveu models, strongly interacting theories) and comparing variational approaches against non-variational scattering protocols.
• Hardware Agnostic: While demonstrated on trapped ions, the method relies on nearest-neighbor and local gates, making it readily adaptable to other architectures like superconducting transmon qubits (e.g., IBM).

In summary, Hite presents a practical, efficient, and experimentally verified method to overcome the depth limitations of NISQ hardware for fermionic scattering, bridging the gap between theoretical quantum field theory simulations and current quantum computing capabilities.