When Google announced quantum supremacy in 2019, the milestone was both real and widely misunderstood. Seven years later, the race has matured — and the definition of “useful quantum advantage” has shifted three times. Here is what actually matters right now.
The state of the machine
Today’s leading superconducting processors hover around 1,000 physical qubits — but physical qubits are noisy. The real race is toward logical qubits, which error-correct thousands of physical qubits down into one reliable computational unit. IBM’s roadmap calls for 200 logical qubits by 2029. The practical implications are already rippling through cryptography and chemistry.
Why error correction changed everything
Surface codes, introduced by Kitaev in 1997, remained theoretical until fabrication caught up. Recent breakthroughs from Google’s Willow chip demonstrated exponential error suppression as code distance grows. That’s the moment the slope of progress went non-linear.
- Physical-to-logical overhead: ~1,000:1 with current codes
- Code distance doubling halves the error rate — demonstrated empirically in 2024
- Concatenated codes + quantum LDPC are the next efficiency frontier
What breaks when this lands
RSA-2048 falls to Shor’s algorithm once you have ~4,000 logical qubits and clean gate operations. That’s not in five years, but it’s not in twenty-five either. Post-quantum cryptography (PQC) migration is now a live enterprise concern — NIST finalized its first set of PQC standards in 2024, and major clouds are already offering PQC key exchange.
The next two years
Watch three things: neutral-atom scaling (Pasqal, QuEra), photonic compilers (PsiQuantum), and the first end-to-end demonstration of a fault-tolerant algorithm on a logical qubit. If any one of these lands, we cross from “exotic research tool” to “compiler target.” That’s the real supremacy moment.