Top 20 Quantum Cloud Computing Terms You Need to Know
The essential vocabulary for accessing quantum power over the internet
You do not need to own a quantum computer to use one. Since IBM opened the first public quantum cloud service in 2016, a rapidly growing ecosystem of platforms has made it possible for anyone with a web browser to design, simulate, and run quantum circuits on real hardware located in data centres around the world. Quantum cloud computing has become the primary way researchers, developers, startups, and enterprises interact with quantum technology, and understanding its vocabulary is essential for navigating this landscape. These 20 terms cover the platforms, tools, access models, and concepts that define quantum computing in the cloud era. For a roundup of available platforms, see 15 Brilliant Platforms To Run Your Quantum Code On.
Quantum Cloud Computing
Quantum cloud computing is the delivery of quantum computing resources over the internet, allowing users to access quantum processors, simulators, and development tools remotely without owning or maintaining quantum hardware. Providers host quantum processing units (QPUs) in specialised facilities with the cryogenic cooling and shielding infrastructure that quantum hardware demands, and expose them to users via APIs, web interfaces, and software development kits. This model has democratised access to quantum technology, enabling a global community of researchers and developers to experiment, prototype, and build quantum applications. For a deeper introduction, see Quantum Cloud Computing: A Guide To A Quantum Future.
Quantum Processing Unit (QPU)
A quantum processing unit is the physical chip or device that performs quantum computations, analogous to a GPU or CPU in classical computing. QPUs vary widely in their underlying technology (superconducting circuits, trapped ions, neutral atoms, photonics) and in their qubit count, connectivity, gate fidelity, and coherence times. In the cloud model, QPUs are housed in provider data centres and accessed remotely. Users typically select a specific QPU backend based on its characteristics and the requirements of their workload.
Quantum as a Service (QaaS)
Quantum as a Service is the commercial model in which quantum computing capabilities are offered on demand over the cloud, typically on a pay-per-use or subscription basis. QaaS encompasses access to QPU hardware, quantum simulators, software development tools, managed notebooks, and consulting services. The model mirrors the evolution of classical cloud computing (IaaS, PaaS, SaaS) and allows organisations to explore quantum computing without capital expenditure on hardware. Major QaaS providers include IBM, Amazon, Microsoft, Google, and a growing number of specialist platforms.
IBM Quantum
IBM Quantum is the pioneering quantum cloud platform, first opened to the public in May 2016. It provides access to a fleet of superconducting quantum processors ranging from utility-scale devices with over 100 qubits to smaller systems for education and prototyping. IBM Quantum is tightly integrated with the open-source Qiskit SDK and offers a web-based composer, managed Jupyter notebooks, and enterprise services through the IBM Quantum Network. It remains the most widely used quantum cloud platform globally. For a walkthrough, see Getting Started With IBM Quantum Cloud.
Amazon Braket
Amazon Braket is the quantum computing service offered by Amazon Web Services (AWS). Unlike single-vendor platforms, Braket acts as a marketplace that provides access to QPUs from multiple hardware providers, including IonQ (trapped ions), Rigetti (superconducting), and QuEra (neutral atoms). Braket integrates with the broader AWS ecosystem, offering managed notebooks, hybrid classical-quantum workflows, and pay-per-shot pricing. Its multi-vendor approach allows users to compare hardware modalities and switch backends without changing their development workflow.
Microsoft Azure Quantum
Azure Quantum is Microsoft’s quantum cloud platform, embedded within the Azure cloud ecosystem. It provides access to QPUs from partners including IonQ, Quantinuum, Rigetti, and Pasqal, alongside Microsoft’s own quantum development tools. Azure Quantum also includes the Azure Quantum Resource Estimator, which calculates the physical resources needed for fault-tolerant algorithms on various architectures. Microsoft’s long-term strategy centres on topological qubits, but Azure Quantum gives users access to today’s leading hardware while the technology matures.
Google Quantum AI
Google Quantum AI is Google’s quantum computing division, responsible for developing its superconducting quantum processors (including Sycamore and Willow) and the open-source Cirq framework. Google provides limited cloud access to its hardware for research partners and selected users. Its 2019 quantum supremacy demonstration and 2024 below-threshold error correction result with the Willow processor are among the most significant milestones in the field. Google’s quantum cloud strategy is closely tied to its broader Google Cloud Platform infrastructure.
Quantum Software Development Kit (SDK)
A quantum SDK is a programming framework that provides the tools for writing, compiling, simulating, and executing quantum circuits. SDKs are the primary interface between developers and quantum cloud platforms. Leading examples include IBM’s Qiskit (Python-based, the most widely adopted), Google’s Cirq, Quantinuum’s TKET, Xanadu’s PennyLane (focused on quantum machine learning), and Amazon’s Braket SDK. Most SDKs support circuit construction, transpilation to specific hardware topologies, noise simulation, and job submission to cloud backends.
Transpilation
Transpilation is the process of converting an abstract quantum circuit into a form that can run on a specific QPU, taking into account the hardware’s native gate set, qubit connectivity, and error characteristics. Because QPUs have limited qubit-to-qubit connections and only support certain gates natively, the transpiler must insert swap operations, decompose gates, and optimise the circuit to minimise depth and error. Transpilation quality directly affects execution fidelity and is a key differentiator among quantum cloud platforms and SDKs.
Quantum Simulator
A quantum simulator (in the cloud context) is a classical computer program that emulates the behaviour of a quantum circuit. Cloud platforms offer simulators that allow users to test and debug their circuits without consuming QPU time, which is more expensive and subject to queue delays. Simulators come in several varieties: statevector simulators that track the full quantum state (limited to around 30-40 qubits), density matrix simulators that model noise, and tensor network simulators that can handle larger circuits with limited entanglement. Most quantum cloud platforms offer simulators as a free or low-cost tier.
Shots
In quantum cloud computing, a shot is a single execution of a quantum circuit followed by measurement. Because quantum measurements are probabilistic, circuits must be run many times (typically hundreds to thousands of shots) to build up a statistical distribution of outcomes. The number of shots directly affects both the accuracy of results and the cost of execution, since most cloud providers price QPU access per shot or per task. Choosing the right number of shots is a practical trade-off between precision and budget.
Job Queue
A job queue is the waiting system that manages access to a QPU when multiple users submit circuits for execution. Because quantum processors are scarce and expensive resources shared among many users, submitted jobs are placed in a queue and executed in order (sometimes with priority tiers based on subscription level). Queue wait times can range from seconds to hours depending on demand and the specific backend. Queue management, fair scheduling, and priority access are important operational aspects of quantum cloud services.
Backend
In quantum cloud terminology, a backend is any execution target to which a user can submit a quantum circuit. Backends include both real QPU hardware and classical simulators. Each backend has a specific set of properties: qubit count, connectivity map, native gate set, calibration data (error rates, coherence times), and current queue status. Cloud platforms typically offer multiple backends, and selecting the right one for a given workload is an important part of the quantum cloud workflow. Backend properties are updated regularly as hardware is recalibrated.
Hybrid Quantum-Classical Workflow
A hybrid quantum-classical workflow is a computation in which a classical computer and a quantum processor collaborate iteratively. The quantum processor handles the parts of the problem where it has an advantage (such as evaluating a quantum state), while the classical computer manages orchestration, data processing, and optimisation loops. Algorithms like VQE and QAOA follow this pattern. Cloud platforms support hybrid workflows through job orchestration tools, classical compute integration, and APIs that allow seamless round-trips between classical and quantum resources.
Quantum Circuit
A quantum circuit is the standard representation of a quantum computation: a sequence of quantum gates, measurements, and resets applied to a register of qubits. In the cloud model, users construct circuits using an SDK, optionally simulate them locally, then submit them to a remote backend for execution. Circuit depth (the number of sequential gate layers), width (the number of qubits), and total gate count are the primary metrics that determine whether a circuit can run successfully on a given QPU, given its coherence time and error rates.
Noise Model
A noise model is a mathematical description of the errors and imperfections present in a specific QPU, including gate error rates, measurement errors, crosstalk between qubits, and decoherence times. Cloud platforms publish calibration data that characterises the noise profile of each backend, and simulators can import these noise models to produce more realistic results. Understanding the noise model is essential for designing circuits that perform well on real hardware, applying error mitigation techniques, and interpreting results correctly.
Quantum Error Mitigation
Quantum error mitigation is a collection of software-level techniques that reduce the impact of hardware noise on quantum results without the full overhead of quantum error correction. Methods include zero-noise extrapolation, probabilistic error cancellation, twirled readout error extinction, and measurement error mitigation. Error mitigation is a standard feature of modern quantum cloud platforms and SDKs, and is essential for extracting useful results from today’s noisy hardware. It represents a practical bridge between the current NISQ era and future fault-tolerant systems.
Quantum Serverless
Quantum serverless is an emerging cloud execution model in which users submit quantum workloads as self-contained programs and the platform handles all resource allocation, scheduling, and execution transparently. It mirrors the serverless paradigm in classical cloud computing (such as AWS Lambda), freeing users from managing backends, queues, and infrastructure. IBM’s Qiskit Runtime and similar offerings are moving toward this model, where users define a quantum program with classical logic and quantum circuits, and the platform orchestrates the entire execution lifecycle.
Blind Quantum Computing
Blind quantum computing is a protocol that allows a client to delegate a quantum computation to a remote server without the server learning anything about the input, the algorithm, or the output. It addresses a fundamental trust problem in quantum cloud computing: when you send your circuit to a provider’s QPU, the provider can in principle see your computation. Blind quantum computing protocols, which require the client to prepare specially encoded qubits, offer privacy guarantees with no classical equivalent. Practical deployment remains a research challenge but is of significant interest for sensitive applications in finance, defence, and healthcare.
Quantum Operating System
A quantum operating system is the software layer that manages the shared resources of a quantum computer in a multi-user cloud environment, handling job scheduling, qubit allocation, calibration cycles, and error correction workflows. As quantum hardware scales and cloud demand grows, efficient resource management becomes critical. Recent research such as the HALO operating system has demonstrated that fine-grained qubit sharing and shot-adaptive scheduling can significantly improve throughput and reduce queue wait times, paving the way for more efficient quantum cloud services.
