Quantum computers are unreliable when noise and errors disrupt the information stored in qubits or the operations performed on them. Errors can arise during state preparation, gates, idle storage and measurement, then accumulate through a circuit. Error mitigation and quantum error correction can improve particular computations, but they do not make every current machine generally fault tolerant. To judge a reliability claim, look at how well a system protects logical information during the target workload—not just its physical-qubit count.
Why can a quantum computer return an unreliable result?
Qubits are sensitive to their surroundings
A qubit’s state can be disturbed by interactions with its environment, producing noise and decoherence. A computation can also be affected by imperfect state preparation, gates, idle memory and measurement. Leakage and other hardware imperfections add further ways for the physical system to depart from the intended computation. Reliability is therefore a property of the whole device and workload, not just one gate’s specification.
As an Amazon Associate I earn from qualifying purchases.
Errors can build up across a circuit
Each operation or period of storage can provide another opportunity for an error to affect the computation. If enough errors accumulate, the output may no longer represent the intended result reliably. The effect depends on the noise and the algorithm: two circuits with similar operation counts need not be equally sensitive.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Does using fewer operations always make a circuit more reliable?
No. Fewer operations can mean fewer opportunities for some errors, but operation count alone does not determine reliability. A 2025 theoretical study by Luis Pedro Garcia-Pintos and coauthors, indexed by NIST, examines coherent, dephasing and depolarizing noise. It cautions that minimizing a compiled circuit’s operation count can be counterproductive if the resulting algorithm is more sensitive to noise under non-ideal conditions. The study offers a framework for reasoning about noise; it is not a benchmark comparing deployed quantum computers.
What is the difference between mitigation, error correction and fault tolerance?
| Approach | What it aims to do | What it does not establish by itself |
|---|---|---|
| Error mitigation | Improve estimates or outputs from a noisy computation using methods suited to a particular setting. | It does not mean errors are detected and corrected as they occur, or that arbitrary long computations will work reliably. |
| Quantum error correction | Encode information across physical qubits as a logical qubit, then use checks for error syndromes to help suppress logical errors. | Encoding does not eliminate the engineering cost: it requires physical qubits, measurements, decoding and control. |
| Fault-tolerant computing | Operate through longer computations while detecting and correcting errors so they do not overwhelm the calculation. | A successful mitigation result or a logical-qubit demonstration does not alone show that a system can reliably run every useful long circuit. |
IBM’s May 2025 explainer defines a fault-tolerant quantum computer as one designed to operate correctly even in the presence of errors. In practice, reaching that goal involves more than the quantum processor: fast measurements and resets, classical decoding, and coordination between the quantum processing unit and classical computing are among the system-level challenges. IBM’s September 2026 discussion describes mitigation and correction as approaches along a path toward fault tolerance and says real-time hierarchical quantum error correction is not directly accessible with current-generation systems.
What does the 2026 logical-qubit demonstration show?
On July 30, 2026, IBM and the University of Chicago announced an encoded-circuit demonstration involving 70 logical qubits, 2,415 logical two-qubit operations and 468 logical T gates. The team reported effective logical error rates 10 times lower than physical error rates. These are results reported by the organizations behind that specific experiment—not a common reliability score for all quantum computers or a cross-platform comparison.
Rank #2
The figures are useful because they describe logical operations and error behavior under encoding, rather than giving only a physical-qubit total. Interpreting them still requires context: which error metric was measured, whether it applies to the entire circuit or a component, how the code and workload were selected, how the result was validated, and what physical and classical resources were used. IBM’s announcement quotes University of Chicago Associate Professor Bill Fefferman saying, “Verification remains one of the biggest challenges in firmly establishing experimental quantum advantage.”
How should you compare claims about quantum-computer reliability?
There is no single field-wide reliability figure established here, and no harmonized current comparison across superconducting, trapped-ion, neutral-atom, photonic and other hardware platforms. A useful comparison needs to keep the device, measurement method, code and workload in view. Check:
- Physical operation quality: gate errors and speeds, with the gate type and measurement method stated.
- Preparation and readout: errors in setting up a state and measuring the result.
- Memory: coherence and idle-storage performance, not only performance while gates are active.
- Connectivity: which qubits can interact directly and how many extra operations are needed to route the workload.
- Logical performance: logical error rates, and whether they improve as the error-correcting code grows or the workload gets longer.
- Correction overhead: physical qubits, measurements, resets and classical decoding resources required per logical operation.
- Workload and verification: the circuit being run, how closely it resembles a useful task, and how the output was checked.
- Evidence status: whether a figure comes from a vendor announcement, a peer-reviewed result or independent replication.
A physical-qubit count or a best-case gate metric cannot answer these questions on its own. Nor should a result for one code or workload be generalized to machines with different architectures and operating conditions.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
