Quantum Hardware Progress: The Race to CRQC

Tracking qubit counts, error rates, and major milestones from leading quantum computing companies

Where We Stand

As of January 2025, the most advanced quantum computers have achieved ~1,000+ physical qubits but only ~5-10 logical qubits (error-corrected, reliable qubits). Breaking cryptocurrency requires ~1,500 logical qubits. At current progress rates, we’re 5-10 years from CRQC, but breakthroughs in error correction could accelerate this timeline significantly.

The Metric That Matters: Logical Qubits

Headlines focus on physical qubit counts (IBM’s 1,121 qubits! Google’s 105 qubits!), but these numbers are misleading. Physical qubits are noisy and error-prone. What matters for breaking cryptography is logical qubits—error-corrected qubits created from many physical qubits.

Metric Physical Qubits Logical Qubits
Definition Raw, noisy qubits Error-corrected, reliable qubits
Current state (2025) 1,000+ achieved ~5-10 achieved
Needed for CRQC ~1-10 million (depends on error rate) ~1,500
Current ratio ~100-1,000 physical qubits per 1 logical qubit

Why the huge overhead? Error correction requires redundancy. To detect and fix errors, you need multiple physical qubits to encode one logical qubit. Better error correction techniques reduce this ratio, which is why error correction breakthroughs are so important.

Leading Quantum Computing Companies

IBM Quantum

Superconducting Qubits

Latest Hardware: Condor (2023)

  • Physical qubits: 1,121
  • Logical qubits: Not yet achieved at scale
  • Error rate: ~10^-3 (0.1% per gate operation)
  • Significance: First to break 1,000 qubit barrier

Roadmap

  • 2025: 4,000+ qubit systems (still physical)
  • 2026-2027: Focus shifts to error correction and logical qubits
  • 2029-2030: Target 100+ logical qubits (IBM’s published goal)

Crypto Implications

IBM is following a “scale up physical qubits first, then improve error correction” strategy. Progress is steady but incremental. If they hit their 2030 target (100 logical qubits), they’ll still be ~15× short of CRQC, but the trajectory would suggest CRQC by 2033-2035.

Google Quantum AI

Superconducting Qubits

Latest Hardware: Willow (2024)

  • Physical qubits: 105
  • Logical qubits: Demonstrated error rates below threshold for quantum error correction
  • Error rate: Below 10^-3, approaching the threshold needed for scalable error correction
  • Significance: Quality over quantity—proves error correction is feasible

Key Achievement

Willow demonstrated that adding more physical qubits for error correction reduces errors rather than amplifying them (a critical milestone called “below threshold”). This proves the path to logical qubits works in practice, not just theory.

Roadmap

  • Google publishes less detailed roadmaps than IBM
  • Focus appears to be on error correction quality first, scaling second
  • If they can maintain below-threshold error rates while scaling, they could leap ahead

Crypto Implications

Google’s approach (perfect the error correction, then scale) could lead to faster progress once they start scaling. Willow’s breakthrough suggests Google might reach 100+ logical qubits sooner than IBM’s timeline. Watch Google closely.

IonQ

Trapped Ion Qubits

Latest Hardware: IonQ Forte (2024)

  • Physical qubits: 32 (smaller than superconducting competitors)
  • Algorithmic qubits: ~29 (IonQ’s metric for “useful qubits”)
  • Error rate: ~10^-4 (better than superconducting qubits)
  • Significance: Higher quality qubits, fewer in number

Trapped Ion Advantages

  • Better connectivity: Any qubit can interact with any other (vs. limited connectivity in superconducting)
  • Lower error rates: Ions are more stable than superconducting circuits
  • Longer coherence times: Can perform more operations before decoherence

Crypto Implications

IonQ’s approach (high-quality, low-quantity) means they might reach logical qubits with fewer physical qubits than IBM or Google. However, scaling to 1,500 logical qubits could take longer. Less immediate threat than superconducting approaches, but don’t discount them.

China (Various Institutions)

Multiple Approaches

Known Progress

  • Physical qubits: 100+ (estimated, some programs classified)
  • Approaches: Superconducting, photonic, trapped ion
  • Funding: Massive government investment (exact figures undisclosed)
  • Transparency: Limited—some achievements published, many likely classified

Key Milestones

  • 2020: Jiuzhang photonic quantum computer claimed quantum advantage
  • 2021: Zuchongzhi superconducting processor (66 qubits)
  • 2023+: Reports of 100+ qubit systems, but details sparse

Crypto Implications

The unknown factor. If China achieves CRQC first, they may not announce it publicly. The NSA likely assumes China’s capabilities are more advanced than publicly disclosed. This is why “store now, decrypt later” attacks are already a concern—adversaries may be collecting encrypted crypto transactions today, betting on quantum decryption within 5-10 years.

The Error Correction Race

The path to CRQC isn’t just about adding more qubits—it’s about reducing the error correction overhead. Current systems need ~100-1,000 physical qubits per logical qubit. If researchers can reduce this to ~10-50 physical qubits per logical qubit, CRQC arrives much faster.

Recent Breakthroughs

  • Surface codes: Most common error correction approach, but heavy overhead
  • Topological codes: Potentially lower overhead, but harder to implement
  • Google’s Willow: Demonstrated exponential error suppression (adding more qubits = fewer errors)
  • New techniques: Researchers exploring alternatives like bosonic codes, cat codes

The Acceleration Risk

Error correction is where breakthroughs happen suddenly. If someone discovers a technique that reduces physical-to-logical overhead from 1000:1 to 50:1, CRQC timelines could compress from 10 years to 3-5 years.

This is why conservative planning (assume 5 years to CRQC, not 10) is prudent for cryptocurrency projects.

What About Other Qubit Technologies?

Technology Status Advantages Challenges
Superconducting Leading (IBM, Google) Fast gates, proven scalability Requires extreme cooling (~0.01K)
Trapped Ion Competitive (IonQ) High fidelity, long coherence Slower gates, scaling challenges
Photonic Early stage Room temperature, networking Difficult to create interactions
Neutral Atom Emerging (Atom Computing) High qubit counts, reconfigurable Lower gate fidelity (improving)
Silicon Spin Research stage Compatible with CMOS, stable Very hard to control, far from scaling

For crypto investors: Superconducting and trapped ion are the near-term threats. Photonic and neutral atom are wildcards—could leap ahead with breakthroughs. Silicon spin is too far out to worry about yet.

Milestones to Watch (2025-2030)

Year Milestone Significance
2025 First 50+ logical qubits demonstrated Proves error correction scales
2026-2027 100+ logical qubits achieved ~7% of the way to CRQC
2028 500+ logical qubits demonstrated ~33% of the way—timeline clarifies
2029 1,000+ logical qubits achieved ~67% of the way—CRQC imminent
2030-2032 1,500+ logical qubits (CRQC) Q-Day arrives

Note: These are projections based on current trends. Breakthroughs or setbacks could shift timelines by ±3-5 years.

Stay Updated on Quantum Progress

Subscribe to our monthly updates tracking qubit milestones, error correction breakthroughs, and what they mean for cryptocurrency security.

Subscribe to Newsletter
Back to Threat Timeline