Abstract. In the pursuit of quantum supremacy, we often forget about practical utility. What if we used a quantum computer not to solve unsolvable problems, but to do useful work faster than classical machines? In this article, we’ll examine the qPoW protocol —quantum mining for blockchain. We’ll explore why a 4-qubit experiment on IBM’s ibmq_quito runs 100 times slower than the simulator, where the break-even point (20–30 qubits) lies, and why CNOT gate noise could bury quantum advantage until the era of fault-tolerant computing.
1. The Stone-Cryptocurrency Paradox
In 2019, the Google team announced they had achieved quantum supremacy: their 53-qubit processor completed a task that would have taken a classical supercomputer 10,000 years. A philosophical parallel arose: “Can Google create a computation so complex that Google itself can’t verify it?” —a variation on the well-known paradox of the omnipotent God and the stone.
“Can Google contrive a computation so complex, Google itself cannot verify it?”
— Figure from the original article (Fig. 1)
The problem is that verifying such calculations requires a classical computer, and if it fails, it’s no longer science, but faith. Researchers from MIT and PoWx have proposed a different approach: Quantum Advantage . This isn’t absolute superiority, but rather a speedup on a specific task that a classical computer can still verify. The ideal candidate is Proof-of-Work (PoW) in blockchain.
2. Mechanism: How does qPoW work?
The protocol is based on the classic SHA3 hashing algorithm. Input data (block number, transaction, nonce) is hashed into a 256-bit string. However, instead of simply trying the nonce (as in classic Bitcoin), qPoW uses a quantum algorithm.
🔹 Coding into rotation angles
The 256-bit string is broken into 4-bit fragments. Each fragment (a number from 0 to 15) is converted into a rotation angle of the quantum gates: Rx , Rz , and cRx .
Example: the fragment 0101 (number 9) gives the angle 9 * π / 8 .
The circuit is measured, and we obtain a 4-bit state, which we consider “most probable.” If the output hash satisfies the difficulty condition (starts with a certain number of zeros), the block has been found.
Cryptographic security: qPoW uses the same architecture as optical PoW, but with a quantum block. Importantly, the result is verified on a classical computer (simulator). If the quantum computer produces an incorrect result due to noise, the block is rejected. This makes the protocol secure and verifiable.
3. Experiment: IBM `ibmq_quito` vs. simulator
The authors ran a 4-qubit circuit on a real public IBM quantum processor and on the Qiskit Aer simulator. The results are shocking without context:
| Parameter | Quantum computer (IBM) | Simulator (qasm) |
|---|---|---|
| Accuracy of state matching | 69% | 100% |
| Execution time (1 iteration) | ~100x slower | Basic |
| Number of CNOT gates (after transpilation) | ~40 (on average) | — |
| CNOT gate error (according to IBM) | ~1% | 0% |
| Readout error | ~2% | 0% |
⚠️ Reality vs. Theory: At 4 qubits, a quantum computer loses to a simulator running on a regular laptop. Why is this necessary? The answer lies in scaling.
Why is the accuracy 69% and not 100%? Transpilation (adapting the circuit to the processor topology) converts logical gates into physical CNOT operations. Each CNOT introduces a ~1% error. 40 CNOTs yield an expected accuracy of (1 – 0.01)^40 ≈ 67% . The experimental value of 69% brilliantly confirms the error theory. But this means that every third mining round will be a waste.
4. Break-even point: 20 qubits
The authors modeled the dependence of execution time on the number of qubits n . For a classical computer, the complexity grows exponentially: t_classical = 10^(0.33n) . For a quantum computer, it grows polynomially: O(n^2) .
The curves intersect at ~20 qubits . Starting at 25–30 qubits, the quantum computer should outperform the simulator. But this is where noise comes into play:
Quantum advantage formula (according to the authors)
Advantage = (Speed_QC / Speed_Classical) × Accuracy
As n increases , the accuracy decreases: (1 – E_CNOT)^CNOTs × (1 – E_readout)^n . If the accuracy drops below ~10%, mining becomes unprofitable, as the circuit would have to be restarted thousands of times. The paper shows that a “minimum viable” advantage requires ~30 qubits, assuming we use native gates and don’t waste time on transpilation.
5. Cryptography and the Future: Risks and Opportunities
qPoW doesn’t threaten Bitcoin’s cryptography (it doesn’t break ECDSA), but rather proposes a change to the consensus mechanism itself. This isn’t an attack , but an evolution: redirecting the energy of mining (which already consumes more electricity than some countries) toward quantum computing R&D.
However, we are faced with open questions from our research constructor:
- Hardware Limitations: Can superconductor technology scale to 30+ qubits with acceptable noise levels? (Modern IBM chips have dozens of qubits, but their connectivity and errors are the main barrier.)
- Economics: Will quantum mining be profitable given the cost of cryogenic installations and error correction?
- Security: Is qPoW vulnerable to classical attacks if an attacker simulates quantum distribution?
6. Conclusion: Lessons for the Researcher
This paper is a fine example of honest science. The authors don’t shout, “We’ve hacked Bitcoin!” They present a 4-qubit prototype that currently performs worse than a classical computer, and openly discuss the conditions under which the coveted quantum advantage will be achieved (20–30 qubits, error suppression, gate optimization) .
What does this mean for us? We are at the dawn of an era when quantum computing is moving from the lab to practical protocols. And while the “quantum stone” is still too heavy for the God-Qubit, we are already learning to lift it.
