Quantum advantage

Quantum advantage is the demonstrated ability of a quantum computer to solve a problem faster or better than the best available classical method. The related but narrower term "quantum supremacy" means only beating classical computers on some task, even a contrived one chosen to be hard classically. Useful advantage, meaning an edge on a problem people actually care about, remains limited as of early 2026.

#Advantage versus supremacy

The word "supremacy" was introduced in 2012 to name the point where a quantum device does something no classical computer feasibly can, regardless of whether the task is useful (Preskill 2012). Many researchers now prefer "quantum advantage," and reserve "practical" or "useful" advantage for the harder goal of outperforming classical methods on a real-world workload. The distinction matters because the first supremacy-style demonstrations used sampling problems engineered specifically to be intractable classically, not tasks of independent value. See Quantum supremacy for the terminology history.

#The 2019 Sycamore claim and the dispute

In 2019 Google reported that its 53-qubit Sycamore processor sampled from a random quantum circuit in about 200 seconds, and estimated the same task would take a leading supercomputer roughly 10,000 years (Arute et al. 2019). This was presented as the first demonstration of quantum supremacy. IBM promptly contested the classical baseline, arguing that with better use of secondary storage a classical supercomputer could perform the task in about 2.5 days rather than millennia (Pednault et al. 2019). Subsequent classical algorithms narrowed the gap further on some instances.

The episode is the standard illustration of why advantage claims are fragile: they compare against the best known classical method, and that baseline can improve after the fact. A demonstration is only as strong as the classical simulation it is measured against, so several sampling-based claims across superconducting and photonic platforms have been partially eroded by improved classical techniques.

#Why useful advantage is still limited

Sampling tasks aside, no quantum computer as of early 2026 has shown a clear advantage on a commercially or scientifically important problem. The reasons are structural:

• Today's machines are in the NISQ era, with noise limiting circuit depth before Quantum error correction can be applied at scale.
• Many proposed near-term algorithms lack proof of advantage, and classical methods, including machine learning and tensor-network simulation, keep advancing.
• Problems with proven exponential quantum speedups, such as factoring via Shor's algorithm, require fault-tolerant machines far beyond current scale.
• Even Quantum annealing hardware, the most deployed by qubit count, has not shown a general practical speedup.

Analysts also distinguish weak advantage, a measurable but modest edge on a narrow task, from strong advantage, an exponential separation on a valuable problem. Only the latter would reshape fields such as chemistry, materials, or cryptanalysis. As of early 2026 the demonstrated results sit at the weak, often contrived, end, and the strong, useful cases remain projected rather than achieved.

#Honest framing

Quantum advantage should be read as a moving, contested benchmark rather than a settled achievement. Narrow supremacy on contrived sampling has been claimed and repeatedly re-examined; broad, useful advantage has not been demonstrated. This is central to the cryptographic picture: the absence of useful advantage today does not imply safety tomorrow, because progress on quantum computers is real, and the harvest now, decrypt later risk means data must be protected before any advantage relevant to cryptography arrives.