Inspiration
Quantum computing opens new frontiers in cryptography, search optimization, and molecular chemistry. This project tackles the three core challenges of the IBM Qiskit Hackathon to benchmark practical quantum algorithms under realistic noise.
What It Does
- E91 QKD: Simulates entanglement-based quantum key distribution using (|\Phi^+\rangle), verifies channel security with the CHSH inequality ((S \approx 2.828)), and accurately classifies eavesdropping attacks versus depolarizing noise.
- Grover's Search under Noise: Evaluates quantum search across various qubit counts ((n \in [2, 10])) under depolarizing and bit-flip noise, mapping the boundaries of quantum advantage over classical baselines.
- VQE for LiH: Implements active-space molecular simulation with (\mathbb{Z}_2) symmetry tapering (reducing 6 qubits to 4) and an ultra-compact 9-CX ansatz, reaching chemical accuracy under strict resource budgets.
Challenges & Solutions
The main challenge was mitigating gate and shot noise in VQE. We designed an ultra-efficient 3-layer ansatz using only 9 CX gates (well within the 40-gate limit) and paired it with a noise-resilient SNR-filtered coordinate optimizer. This achieved fast, reliable convergence (~2.4s per geometry) while staying well inside the 20M shot ceiling.
What We Learned
We gained hands-on experience in error budget management, symmetry reduction techniques, and balancing circuit depth against noise in NISQ-era quantum computing with Qiskit.
