The State of Quantum Computing Knowledge

The State of Quantum Computing Knowledge

Abstract

This document provides a comprehensive survey of the current state of quantum computing: what we know, what works, what doesn't, and the fundamental challenges that remain. Understanding these foundations is essential for identifying where AI can most effectively accelerate quantum hardware development.

1. Hardware Paradigms

1.1 Technology Landscape (2024-2026)

1.2 Superconducting Qubits

Physical Implementation:

• Transmon qubits: Josephson junction-based anharmonic oscillators
• Operated at 10-20 mK (dilution refrigerator)
• Microwave control pulses (4-8 GHz)

Advantages:

• Mature fabrication (semiconductor processes)
• Fast gates (10-100 ns)
• Scalable manufacturing

Challenges:

• Short coherence times
• Requires extreme cooling
• Crosstalk between neighboring qubits

Key Metrics (IBM Heron, 2024):

T1 (relaxation):     300 μs
T2 (dephasing):      200 μs
Single-qubit gate:   99.9%
Two-qubit gate:      99.5%
Readout fidelity:    99.5%

1.3 Trapped Ion Qubits

Physical Implementation:

• Individual ions (Yb+, Ca+, Ba+) trapped in electromagnetic fields
• Laser-cooled to ground state
• Laser pulses for gates

Advantages:

• Longest coherence times
• Highest gate fidelities
• All-to-all connectivity

Challenges:

• Slower gates (100 μs - 1 ms)
• Scaling beyond ~50 qubits difficult
• Complex laser systems

Key Metrics (Quantinuum H2, 2024):

T1:                  >10 seconds
T2:                  ~1 second
Single-qubit gate:   99.99%
Two-qubit gate:      99.7%
Readout fidelity:    99.9%

1.4 Photonic Qubits

Physical Implementation:

• Photons as qubits (polarization, path, time-bin encoding)
• Integrated photonic circuits
• Single-photon sources and detectors

Advantages:

• Room temperature operation
• Natural for quantum communication
• Low decoherence

Challenges:

• Probabilistic gates
• Single-photon generation difficult
• No native qubit-qubit interaction

Key Developments:

• Xanadu Borealis: 216 squeezed-state modes
• Gaussian Boson Sampling demonstrated
• Measurement-based quantum computing approach

1.5 Neutral Atom Qubits

Physical Implementation:

• Individual atoms (Rb, Cs) trapped in optical tweezers
• Rydberg states for interactions
• Reconfigurable array geometry

Advantages:

• Identical qubits (no fabrication variation)
• Scalable to 1000+ atoms
• Reconfigurable connectivity

Challenges:

• Slower gates than superconducting
• Atom loss during computation
• Rydberg state lifetime limits

Key Metrics (QuEra Aquila, 2024):

Atoms:              256
Coherence:          ~1 second
Two-qubit gate:     99.5%
Array geometry:     Fully programmable

2. Proven Quantum Advantages

2.1 Shor's Algorithm (Factoring)

Classical complexity: O(exp(n^(1/3) · (log n)^(2/3))) - sub-exponential but impractical

Quantum complexity: O((log N)³) - polynomial time

Status:

• Proven correct
• Demonstrated on small numbers (15, 21)
• Requires fault-tolerant hardware for cryptographic relevance
• RSA-2048: Requires ~4000 logical qubits, millions of physical qubits

Mathematical Foundation:

Given N = p × q (product of two primes)
1. Choose random a < N
2. Find period r of f(x) = a^x mod N using QFT
3. Factor N using r: gcd(a^(r/2) ± 1, N)

Quantum speedup: Period finding uses O(log N) qubits
                 and O((log N)³) gates

2.2 Grover's Algorithm (Unstructured Search)

Classical complexity: O(N) - must check each item

Quantum complexity: O(√N) - quadratic speedup

Status:

• Proven optimal for unstructured search
• Demonstrated on current hardware (small N)
• Useful for cryptography, optimization, database search

Mathematical Foundation:

Given oracle O: O|x⟩ = (-1)^f(x)|x⟩ where f(x) = 1 for target

Grover iteration:
1. Apply oracle O (mark target)
2. Apply diffusion D = 2|ψ⟩⟨ψ| - I (amplify)

After k = π/4 · √(N/M) iterations:
P(measuring target) = sin²((2k+1)θ) ≈ 1

where sin²(θ) = M/N, M = number of targets

2.3 Quantum Simulation

Problem: Simulate quantum systems (molecules, materials)

Classical complexity: O(2^N) - exponential in particle count

Quantum complexity: O(N^4) - polynomial for many systems

Status:

• Native advantage (quantum simulating quantum)
• VQE demonstrated for small molecules
• Error mitigation critical for NISQ devices

Applications:

• Drug discovery (protein folding, binding)
• Materials science (superconductors, catalysts)
• Chemistry (reaction mechanisms)

Key Formula (Hamiltonian Simulation):

H = Σᵢ hᵢ Pᵢ  (Pauli decomposition)

U = e^(-iHt) ≈ (Πᵢ e^(-ihᵢPᵢt/r))^r  (Trotter-Suzuki)

Quantum advantage: Represent 2^N amplitudes with N qubits

2.4 Computational Supremacy Demonstrations

Google Sycamore (2019):

• 53 superconducting qubits
• Random circuit sampling
• Claimed: 200 seconds vs 10,000 years classical
• Contested: Classical algorithms improved

USTC Jiuzhang (2020):

• 76 photons, 100 modes
• Gaussian Boson Sampling
• Claimed: 200 seconds vs 600 million years
• Specific to sampling problem

IBM Response:

• Showed classical simulation in ~2.5 days
• Debate ongoing on "useful" supremacy

3. The NISQ Era

3.1 Definition (Preskill, 2018)

NISQ = Noisy Intermediate-Scale Quantum

Characteristics:

• 50-1000 qubits
• Limited coherence (microseconds to seconds)
• No quantum error correction
• Gate errors ~0.1-1%
• Shallow circuit depth

3.2 Error Budget Analysis

For a circuit with depth d and n qubits:

Total error ≈ 1 - (1 - ε_1q)^(n×d) × (1 - ε_2q)^(n_2q) × (1 - ε_r)^n

Example (100 qubits, depth 100):
ε_1q = 0.001, ε_2q = 0.01, ε_r = 0.01
P(success) ≈ exp(-100×100×0.001 - 50×100×0.01 - 100×0.01)
           ≈ exp(-10 - 50 - 1)
           ≈ exp(-61)
           ≈ 0  (essentially zero)

Implication: NISQ circuits limited to depth ~10-100 with current error rates.

3.3 Error Mitigation Strategies

3.4 Variational Algorithms (NISQ-native)

VQE (Variational Quantum Eigensolver):

min_θ ⟨ψ(θ)|H|ψ(θ)⟩

1. Prepare ansatz |ψ(θ)⟩ with parameterized gates
2. Measure expectation value of Hamiltonian terms
3. Classical optimizer updates θ
4. Repeat until convergence

QAOA (Quantum Approximate Optimization Algorithm):

|ψ(γ,β)⟩ = e^(-iβₚB) e^(-iγₚC) ... e^(-iβ₁B) e^(-iγ₁C) |+⟩^n

C = cost Hamiltonian (encodes optimization problem)
B = mixer Hamiltonian (enables exploration)

Challenges:

• Barren plateaus (vanishing gradients)
• Local minima
• Hardware noise corrupts gradients
• Limited expressivity vs depth

4. Fundamental Limits

4.1 Threshold Theorem

Statement: If the error rate per gate is below a threshold ε_th, arbitrarily long quantum computations are possible with polynomial overhead.

Thresholds:

Current Hardware:

• Best 2-qubit gates: ~0.3% error
• We are at or near threshold for surface codes

4.2 Error Correction Overhead

For surface code with physical error rate p:

Logical error rate: p_L ≈ 0.03 × (p/p_th)^((d+1)/2)

where d = code distance

Example: Achieve p_L = 10⁻¹⁵ (useful for Shor's algorithm)
         with p = 0.1%, p_th = 1%, d = 27

Physical qubits per logical qubit: 2d² ≈ 1458
For 4000 logical qubits: ~6 million physical qubits

4.3 No-Cloning Theorem

Statement: It is impossible to create an identical copy of an arbitrary unknown quantum state.

Proof:

Assume cloning unitary U exists:
U|ψ⟩|0⟩ = |ψ⟩|ψ⟩
U|φ⟩|0⟩ = |φ⟩|φ⟩

Then: ⟨ψ|φ⟩ = (⟨ψ|⟨0|)(|φ⟩|0⟩) = (⟨ψ|⟨ψ|)(|φ⟩|φ⟩) = (⟨ψ|φ⟩)²

This is only satisfied if ⟨ψ|φ⟩ = 0 or 1 (orthogonal or identical)
Contradiction for arbitrary states.

Implications:

• Cannot amplify quantum signals classically
• Quantum error correction must encode, not copy
• Enables quantum cryptography (eavesdropping detection)

4.4 Holevo Bound

Statement: N qubits can transmit at most N classical bits of accessible information.

χ(ρ) ≤ S(ρ)  (von Neumann entropy)

where χ = Σᵢ pᵢ S(ρᵢ) - S(Σᵢ pᵢ ρᵢ)

Implication: Quantum advantage comes from computation, not storage capacity.

5. Open Problems

5.1 Fault-Tolerant Quantum Computing at Scale

Challenge: Build a quantum computer with enough logical qubits for useful algorithms.

Requirements:

• Millions of physical qubits
• Error rates below threshold
• Fast classical decoding
• Real-time error correction

Progress:

• Google: "Below threshold" demonstration (2024)
• IBM: 1000+ qubit systems
• Logical qubit demonstrations: 1-10 logical qubits

5.2 Quantum Memory (QRAM)

Challenge: Random access to quantum data in O(log N) time.

Bucket-Brigade Proposal:

|i⟩|0⟩ → |i⟩|Dᵢ⟩

Uses binary tree of routers
O(log N) qubits for address
O(N) memory cells

Status:

• Theoretical proposals exist
• No practical implementation
• Critical for many quantum ML algorithms

5.3 Practical Quantum Advantage

Challenge: Demonstrate quantum advantage for economically useful problems.

Candidates:

• Quantum chemistry (drug discovery)
• Optimization (logistics, finance)
• Machine learning (specific problems)

Barriers:

• Error rates too high for useful depth
• Classical algorithms keep improving
• Problem encoding overhead

5.4 Room-Temperature Qubits

Challenge: Build qubits that operate without extreme cooling or isolation.

Approaches:

• NV centers in diamond (optical)
• Topological qubits (if they exist)
• Novel materials

Status: No viable path currently

6. Key Metrics Summary

6.1 Where We Are

6.2 Improvement Trajectory

Year    Qubits    2Q Error    Logical Qubits
2019    53        0.6%        0
2021    127       0.5%        0
2023    1000+     0.3%        1-2
2025    4000+     0.1%?       10-100?
2030    >10000    <0.01%?     1000+?

7. Conclusion: Where AI Can Help

The analysis above reveals clear bottlenecks where AI can accelerate progress:

These opportunities are explored in detail in the subsequent documents.

References

1. Nielsen, M. A., & Chuang, I. L. (2010). Quantum Computation and Quantum Information. Cambridge University Press.

2. Preskill, J. (2018). "Quantum Computing in the NISQ era and beyond." Quantum, 2, 79.

3. Arute, F., et al. (2019). "Quantum supremacy using a programmable superconducting processor." Nature, 574(7779), 505-510.

4. Zhong, H. S., et al. (2020). "Quantum computational advantage using photons." Science, 370(6523), 1460-1463.

5. Acharya, R., et al. (2024). "Quantum error correction below the surface code threshold." Nature, 634, 315-320.

6. Gottesman, D. (2010). "An introduction to quantum error correction and fault-tolerant quantum computation." Quantum Information Science and Its Contributions to Mathematics, 68, 13-58.

7. Giovannetti, V., Lloyd, S., & Maccone, L. (2008). "Quantum random access memory." Physical Review Letters, 100(16), 160501.

8. McClean, J. R., et al. (2016). "The theory of variational hybrid quantum-classical algorithms." New Journal of Physics, 18(2), 023023.

Document: 01_quantum_state_of_knowledge.md
Part of: AI-Quantum Symbiosis Research Series
Location: Documents/Learning/ai_quantum_symbiosis/