IBM Quantum

IBM Quantum is IBM's quantum-computing division, which builds gate-model quantum computers from superconducting transmon qubits and develops the open-source Qiskit software stack. It runs one of the largest fleets of cloud-accessible quantum processors and publishes a detailed hardware roadmap that it revises over time.

#Processors

IBM names its processors after birds and has scaled qubit counts steadily. The Eagle processor reached 127 qubits in 2021, Osprey reached 433 qubits in 2022, and Condor reached 1121 qubits in 2023, at the time the largest single superconducting chip announced. IBM then shifted emphasis from raw qubit count to quality, introducing the Heron family, a modular processor with tunable couplers and substantially lower error rates than the earlier monolithic chips. As of early 2026, IBM's strategy centers on interconnecting Heron-class chips rather than building ever-larger single dies, a change reflected in its published roadmap. Heron introduced tunable couplers, which let the machine turn interactions between neighboring qubits on and off, reducing the crosstalk errors that limited the earlier fixed-coupling designs.

IBM also operates the IBM Quantum Network, through which companies, national laboratories, and universities access its processors over the cloud, and it has published a series of quantum systems since the first cloud-accessible device in 2016.

#IBM Quantum System Two

IBM Quantum System Two is a modular hardware platform designed to house multiple processors and the cryogenic and control infrastructure they need. It is the system IBM positions as the foundation for scaling toward larger machines by linking chips together, in contrast to the single-chip System One that preceded it.

#Qiskit and Quantum Volume

Qiskit is IBM's open-source SDK for building, compiling, and running quantum circuits, widely used in research and teaching. IBM also introduced Quantum Volume, a single-number benchmark that accounts for qubit count, connectivity, gate fidelity, and other error sources by measuring the largest square random circuit a machine can run reliably (Cross et al. 2019). Quantum Volume was designed so that adding noisy qubits does not inflate the score, making it a more honest measure than qubit count alone. It has since been complemented by speed and application-oriented benchmarks across the field.

#Utility experiments

In 2023 IBM reported a "utility" experiment on a 127-qubit Eagle processor, using error mitigation to compute properties of a quantum spin model at a scale where brute-force classical simulation is difficult (Kim et al. 2023). Several groups responded with classical methods, including tensor-network and other approximations, that reproduced or improved on the results, so the experiment is best understood as a demonstration that pre-fault-tolerant machines can produce useful physics estimates rather than as proof of a decisive advantage. IBM framed the work as evidence of utility before fault tolerance, not as a supremacy claim.

#Roadmap toward error correction

IBM's public roadmap sets targets, not achieved results, and the company states this explicitly. The near-term milestones emphasize Quantum error correction and modular scaling. IBM has described a system named Starling as a target for a large-scale, error-corrected machine around 2029, with a stated goal on the order of 200 logical qubits (IBM roadmap). These figures are planning targets. Every published quantum roadmap in the field, IBM's included, should be read as an aspiration subject to revision, and IBM has revised its own roadmap more than once. As of early 2026 no vendor, IBM included, has demonstrated a large fault-tolerant machine.

#Relation to cryptography

IBM's processors run quantum logic gates and are in principle universal, so a sufficiently large, error-corrected version could run Shor's algorithm. Current Heron-class machines are far from that scale. IBM has also been active in Post-quantum cryptography, contributing to standardized algorithms, so its work spans both the hardware that motivates the quantum threat and the classical cryptography meant to withstand it.