D-Wave Annealing vs Gate-Based Quantum Computers: Why 5,000 Qubits Don't Break RSA
Two Fundamentally Different Machines
D-Wave (Burnaby, BC, Canada) has built the largest quantum systems by raw qubit count. Its Advantage processor has over 5,000 qubits with the Pegasus topology (15-way connectivity, 35,000 connections, over 1 million Josephson junctions). This is more physical qubits than IBM Condor (1,121), Google Willow (105), Zuchongzhi 3.0 (105), and Atom Computing (1,180) combined.
Yet D-Wave cannot break RSA. That is because D-Wave is a quantum annealer, not a gate-based universal quantum computer.
Annealing vs Gate-Based
Gate-based quantum computing (IBM, Google, IonQ, Quantinuum) executes arbitrary quantum algorithms by composing sequences of quantum gates. Shor's algorithm is expressible as a gate sequence. Grover's algorithm is expressible as a gate sequence. Universal gate-based quantum computing can solve any problem a quantum Turing machine can solve.
Quantum annealing (D-Wave) solves optimization problems by encoding them as Ising spin models and letting the system evolve to a low-energy state under a Hamiltonian that's slowly changed. It is good at specific optimization problems. It cannot run arbitrary quantum algorithms.
Why Annealing Does Not Break Encryption
Shor's algorithm uses the quantum Fourier transform, which requires coherent gate operations across the entire computational register. A quantum annealer cannot implement arbitrary gate sequences. There is no known mapping of Shor's algorithm onto quantum annealing. D-Wave's 5,000+ qubits cannot factor RSA-2048 no matter how large the machine becomes.
Grover's algorithm has a similar story. While some variants of quantum search can be reformulated as annealing problems, the speedup over classical search is much weaker than Grover's O(sqrt(N)).
What Annealing Can Do
Annealing is genuinely useful for:
- Combinatorial optimization (traveling salesman, portfolio optimization, scheduling).
- Machine learning (training Boltzmann machines).
- Simulation of Ising spin systems.
- Some quadratic unconstrained binary optimization (QUBO) problems.
D-Wave customers include Volkswagen (route optimization), NASA (task scheduling), and banks (portfolio optimization).
The Qubit-Count Marketing Confusion
Because D-Wave reports qubit counts in the thousands while IBM and Google report in the hundreds, a casual reader can conclude D-Wave is "ahead." This is wrong in every way that matters for cryptography. The correct comparison is apples to oranges: D-Wave is a special-purpose optimization machine, IBM/Google are general-purpose computers at an earlier scale.
D-Wave's Fifth-Generation System
D-Wave's Advantage 2 prototype continues the annealer line with improved coherence and connectivity. D-Wave is public (NYSE: QBTS) and one of three publicly traded pure-play quantum companies.
Implications
- Do not treat D-Wave qubit counts as CRQC signals. They are not comparable to gate-based counts.
- Annealing for optimization is real but off-cryptographic-path. It does not accelerate the CRQC timeline.
- PQC migration urgency is driven by gate-based progress, not by D-Wave.
- D-Wave's annealing progress is still valuable for specific optimization workloads, unrelated to PQC.
Frequently Asked Questions
Can D-Wave break RSA?
No. D-Wave is a quantum annealer that solves Ising spin optimization problems. It cannot run Shor's algorithm, which requires gate-based quantum computation. No D-Wave system can break RSA.
How many qubits does D-Wave Advantage have?
Over 5,000 qubits with the Pegasus topology, 15-way connectivity, 35,000 connections, and over 1 million Josephson junctions. A much larger number than any gate-based system but not comparable.
Is D-Wave useful for anything?
Yes. D-Wave annealers are useful for certain combinatorial optimization, machine learning training, and Ising spin simulation. Customers include Volkswagen, NASA, and financial services firms.
Should I worry about D-Wave's growing qubit count?
For cryptography, no. Annealer qubits cannot run Shor. For optimization workloads you care about, possibly yes; D-Wave's annealing may offer practical advantage on specific problem classes.
Sources
Related Articles
Protect Your Data Before Q-Day Arrives
QNSQY's NIST-standardized post-quantum encryption protects files against both current and quantum-era threats.
Try QNSQYOriginally published at quantumsequrity.com/blog/dwave-annealing-vs-gate.