← Back to ranked candidates
Data·Composite 3.30 / 5 (default weights)

Quantum Topological Data Analysis

Exponential speedup for Betti numbers — if dequantization fails

Thesis

QTDA — computing the Betti numbers of a simplicial complex via quantum projection onto the kernel of a combinatorial Laplacian — was the canonical example of an exponential-speedup quantum algorithm for a 'natural' data-analysis task. Lloyd–Garnerone–Zanardi (2014) and follow-ups give an O(poly log N) routine for a problem whose best classical algorithms scale polynomially in N. The question of whether this represents a true Shor-level breakthrough hinges entirely on the dequantization frontier: Apers–Gribling, Gharibian–Le Gall and others have classical algorithms with overlapping regimes, while Quantinuum's 2023–2024 H-series demonstrations show end-to-end QTDA on real hardware. The most likely outcome is a 'conditional Shor': a real exponential separation on data structures with high Betti-to-simplex ratio that classical sampling cannot exploit.

Scoring

Provable speedup4/5
Dequantization resilience2/5
Resource efficiency4/5
Application breadth3/5
Empirical traction3/5

Genuine candidate for end-to-end exponential separation, but threatened by ongoing classical sampling improvements. Resource-cost-favourable (small qubit counts work meaningfully), and the only family in this dashboard with a working hardware demo at modest scale.

Open problems

  • ·Identifying datasets with provably high Betti-to-simplex ratio resistant to classical sampling.
  • ·Closing the gap between Lloyd-style heuristic speedup and Berry-style provable speedup.
  • ·Block-encoding the boundary operator efficiently for sparse complexes.
  • ·QTDA on persistence (filtration of complexes), not just a single complex.
  • ·Hardware demonstrations beyond ~20-simplex toys.

Demo · pre-computed on the Selene emulator

QTDA — SWAP-test fidelity matrix and Vietoris-Rips graph

1.00
0.99
0.40
0.27
0.99
1.00
0.45
0.36
0.39
0.46
1.00
0.99
0.33
0.36
0.98
1.00
Patients
4
Pairs
6
Threshold
0.85
β₀ (components)
2

Each cell is the inner-product fidelity between two amplitude-encoded patient states, estimated from 1000 Selene shots per pair. Edges form when fidelity ≥ threshold; Betti-0 counts connected components. A canonical hardware demonstration ran a 28-pair version on the H-series (Akhalwaya et al. 2023).

Key papers (8)

Quantum algorithms for topological and geometric analysis of data
Lloyd, Garnerone, Zanardi · 2014
arXiv:1408.3106

The original QTDA paper. Claims exponential speedup for Betti number estimation over the best then-known classical algorithms.

Quantum algorithms for topological data analysis with provable speedups
Berry, Su, Casares, McArdle, Tomesh · 2022
arXiv:2209.12887

First QTDA algorithm with rigorous, end-to-end speedup analysis — clarifies the regime where quantum is genuinely faster.

Dequantizing the Quantum Singular Value Transformation: Hardness and Applications
Bakshi, Tang · 2023
arXiv:2303.01492

Provides classical algorithms matching QSVT-based QTDA under sampling access — bounds where QTDA's exponential edge actually survives.

Towards Quantum Advantage on Noisy Devices with Instance-Specific Speedups
McArdle, Gilyén, Berta · 2022
arXiv:2206.02639

Resource-estimates QTDA on near-term and early-FT devices; identifies instance families with concrete quantum advantage.

Towards Quantum Advantage via Topological Data Analysis
Schmidhuber, Lloyd · 2022
arXiv:2209.13581

Argues that high-Betti-number regimes resist all known classical sampling attacks — likely home of any future QTDA quantum advantage.

Complexity-Theoretic Limitations on Quantum Algorithms for TDA
Crichigno, Kohler · 2022
arXiv:2209.14286

Shows that Betti-number estimation in the worst case is #P-hard, so any exponential speedup must come from structure — not from the problem class alone.

End-to-end Resource Analysis for Quantum Interior Point Methods and Portfolio Optimization
Dalzell, Clader, Salton, et al. · 2022
arXiv:2211.12489

Methodologically representative end-to-end FT resource analysis applied to data problems — the same lens being used for QTDA.

Topological Data Analysis on Noisy Quantum Computers
Akhalwaya, Ubaru, Clarkson et al. (IBM Research) · 2022
arXiv:2209.09371

End-to-end NISQ QTDA demonstration on IBM hardware — small but real evidence that the pipeline runs. Accepted at ICLR 2024.