That does not mean quantum computers can break encryption tomorrow. The hardware still has a long way to go. But it does mean that governments, businesses and critical infrastructure providers should take post-quantum security seriously today.
Because the risk now lies in what can be collected and decrypted later.
Why this is important for businesses and society
Encryption protects much of modern digital life: banking, medical records, government systems, business contracts, intellectual property, cryptocurrencies and secure communications. Many of these systems rely on cryptographic methods such as ECC and RSA. These are strong against classical computers, but they are known to be vulnerable to sufficiently powerful quantum computers.
Sensitive data often needs to remain confidential for years or even decades. If encrypted data is copied now, it could potentially be stored and decrypted in the future when quantum computers become powerful enough. This is often called a “harvest now, decrypt later” risk. Post-quantum readiness is therefore a trust, resilience and business continuity issue.
Three papers, one message
Over the past few weeks, three research papers have shown progress in reducing the estimated resources needed for a cryptographically relevant quantum computer.
Each paper approaches the challenge differently. Together, they show how quickly the field is improving, not only through better hardware, but also through smarter algorithms, error correction and system design.
Google Quantum AI: faster attacks with fewer resources
On 31 March, Google Quantum AI published a zero-knowledge proof showing that it had designed a quantum circuit capable of solving the elliptic curve discrete logarithm problem for ECDLP 256 on the secp256k1 curve. This is the elliptic curve used in several cryptocurrency systems.
According to the paper, the proposed design would require around 500,000 physical qubits and could complete the computation in approximately nine minutes on a superconducting quantum platform. That represents a major improvement in estimated spacetime volume compared to previous approaches.
Google did not disclose the full circuit. Instead, the team published a zero-knowledge proof: a cryptographic way of proving that their circuit achieves the claimed performance without revealing the circuit itself. This is an interesting development in its own right. It shows how cryptographic proof techniques can be used to validate sensitive quantum research claims while keeping key details private.
Oratomic: fewer qubits, longer runtime
On the same day, Oratomic qnd Q-CTRL released a paper exploring the same ECC scheme using a neutral atom quantum platform. Their results suggest that the computation could theoretically be performed with around 10,000 physical qubits, but with a much longer runtime of approximately 264 days. A second configuration using around 26,000 physical qubits could reduce the runtime to roughly ten days.
This is a different trade-off. Neutral atom systems are generally slower than superconducting systems, but they may offer advantages in scalability and coherence. In other words, they may be able to keep quantum information stable for longer. A computation that takes months may not sound like an immediate threat. But for stored data, speed is not always the most important factor. If an attacker is targeting data at rest, such as archived documents, copied databases or long-term confidential files, a slower attack may still be relevant. However, there is an important caveat. Oratomic’s results rely on qLDPC codes for error correction. These codes can significantly reduce qubit requirements, but they are less mature and less well understood than the surface code error correction used in Google’s proposed approach.
Q-CTRL: making better use of quantum hardware
A third paper, published on 7 April by Q-CTRL, takes a different angle. Instead of introducing a new algorithm, it focuses on improving how quantum hardware is used. The key observation is that many qubits remain idle during parts of a computation. By making better use of those idle periods, the overall resource requirements could be reduced.
The proposed architecture takes inspiration from classical computing. It separates the system into two parts:
- Quantum processing
A fast quantum processor, such as a superconducting platform, performs operations quickly. - Quantum memory
A separate quantum memory system, potentially based on technologies with longer coherence times such as neutral atoms or trapped ions, stores quantum information more efficiently.
This hybrid approach could reduce the need for error correction and lower resource requirements by an additional factor of two to five.
A new divide is emerging in quantum computing
Taken together, these papers show that the quantum computing scene is becoming more specialised. On one side, we see fast platforms such as superconducting systems. These may be able to execute attacks quickly, potentially within minutes, once they reach sufficient scale and reliability. On the other side, we see slower platforms with longer coherence times, such as neutral atoms or trapped ions. These may be better suited to long-running computations or attacks on stored data, where immediate speed is less important.
This distinction matters for cybersecurity. Fast platforms could create risks for systems that need immediate protection, such as certain cryptocurrency transactions. Slower platforms could still pose a serious threat to archived or long-term sensitive data.
But the most important takeaway is not that quantum computers can already break encryption. They cannot. The important takeaway is that the estimated threshold keeps moving. Resource requirements are decreasing because researchers are improving quantum circuits, error correction, hardware utilisation and system architecture. Each improvement brings the field closer to the point where quantum computers become relevant for actual cryptographic risk.
What organisations should do now
For businesses, governments and infrastructure providers, the next step is to prepare. Organisations should start by understanding where they use vulnerable cryptography, which data needs long-term protection and which systems may be difficult to migrate later. Post-quantum migration requires planning, testing, coordination and ecosystem-wide collaboration.
The research from Google Quantum AI, Oratomic and Q-CTRL shows that the road to cryptographically relevant quantum computing may be shorter than many expected. The hardware challenges remain significant, but the direction is clear.
Disclaimer
These papers reduce the estimated theoretical resource requirements for breaking certain cryptographic schemes. They do not mean that such attacks are possible today.
Significant breakthroughs are still needed in quantum hardware fabrication, scalability, error correction and reliability before the architectures described in these works can be realised in practice.
