HWCO-QPU · REV AQPU intelligence · Quantum hardware

Every quantum computer, on one sheet.

Compare quantum processors by architecture, qubits, fidelity, provider and access, from published specifications with every figure sourced. Then find the QPU that suits your workload.

Explore QPUs →Compare processors

Fig. 01 · Quantum metrology lab, NASA Glenn2023

QPUs Tracked

Physical quantum processors in database

Providers

Hardware manufacturers & cloud operators

Architectures

Distinct qubit modalities covered

Cloud Platforms

Access platforms with QPU availability

Benchmarks

Published performance measurements indexed

Sample data — specifications updated as verified. See methodology for sourcing details.

QPU DATABASE

Explore Quantum Processors

View all QPUs →

Processor	Provider	Architecture	Qubits	Connectivity	Access	Status
IBM Eagle r1	ibm	Superconducting	127	heavy-hex	Cloud	Cloud Access
IBM Heron r2	ibm	Superconducting	156	heavy-hex	Cloud	Cloud Access
IonQ Forte	ionq	Trapped Ion	32	all-to-all	Cloud	Cloud Access
IonQ Forte Enterprise	ionq	Trapped Ion	35	all-to-all	Cloud	Private Access
Rigetti Ankaa-3	rigetti	Superconducting	84	square lattice	Cloud	Cloud Access
Quantinuum H2-1	quantinuum	Trapped Ion	56	all-to-all	Cloud	Reservation
QuEra Aquila	quera	Neutral Atom	256	reconfigurable 2D array	Cloud	Cloud Access
IQM Spark	iqm	Superconducting	5	ring	Cloud	Cloud Access

View All Quantum Processors →

HARDWARE MODALITIES

Quantum Computers Aren't All the Same.

QPU architecture fundamentally determines connectivity, gate operations, coherence characteristics, operating environment, and workload suitability.

Superconducting

›Fast nanosecond gate operations
›Mature fabrication ecosystem
›Fixed nearest-neighbor connectivity

→ Learn more

Trapped Ion

›Highest gate fidelities (~99.9%)
›Native all-to-all connectivity
›Coherence times measured in seconds

→ Learn more

Neutral Atom

›Reconfigurable qubit connectivity
›Large arrays (100–1000+ atoms)
›Natural analog simulation capability

→ Learn more

Photonic

›Room-temperature operation feasible
›Fiber-compatible — quantum networking
›No decoherence from thermal noise

→ Learn more

Quantum Annealing

›Largest qubit counts commercially (5000+)
›Native optimization via Ising model
›Hybrid classical-quantum solvers

→ Learn more

Topological

›Theoretically intrinsic fault tolerance
›Non-local qubit encoding
›Research stage — not yet commercial

→ Learn more

COMPARISON TOOL

Compare QPUs Side by Side

Add any two or three processors to compare architecture, qubit count, connectivity, fidelity, access options, and pricing in one view.

IONQ

IonQ Forte

Trapped Ion

ArchitectureTrapped Ion

Physical Qubits32 qubits

ConnectivityAll-to-all

Gate Fidelity99.5%

IBM

IBM Heron r2

Superconducting

ArchitectureSuperconducting

Physical Qubits156 qubits

Connectivityheavy-hex

Gate FidelityPublished

Add ProcessorSlot 3

Open QPU Comparison →

Comparison Dimensions

Architecture

Physical Qubits

Connectivity

Gate Fidelity

Access Platforms

Pricing

Framework Support

QPU ADVISOR

Tell Us the Problem.

We'll Help Find the Hardware.

Advisor OutputEXAMPLE

Recommended Architecture

Trapped Ion

Best Matches

4 QPUs

Confidence

High

Reasoning, access paths, and framework compatibility appear here after workload analysis.

CLOUD ACCESS LAYER

One Hardware Market. Multiple Ways to Access It.

Quantum processors are accessible through major cloud platforms, direct provider portals, and on-premises deployments. hardware.co tracks availability across all access routes.

Amazon Braket

Providers

IonQRigettiQuEra

Frameworks

QiskitBraket SDKPennyLane

Multi-region

View Access Options →

Azure Quantum

Providers

IonQQuantinuumRigetti

Frameworks

Q#QiskitCirq

US East · EU West

View Access Options →

IBM Quantum

Providers

IBM

Frameworks

QiskitPennyLaneCUDA-Q

US · EU · JP

View Access Options →

Direct Provider Access

Access hardware directly from the manufacturer via dedicated cloud portals, on-premises deployment, or enterprise agreements.

Providers

IonQQuantinuumD-WavePASQALIQM

Frameworks

Provider-specific

Varies by provider

View Access Options →

WORKLOAD MATCHING

Matching Workloads to Architecture

Different quantum architectures excel at different problem types. Use this reference to identify which hardware modalities best fit your workload category.

Workload	Superconducting	Trapped Ion	Neutral Atom	Photonic	Annealing
Chemistry
Materials
Optimization
Finance
ML
Drug Discovery
Logistics

Legend:StrongPossibleExperimentalLimited

Conceptual suitability assessment for current NISQ-era hardware. Actual performance varies. See methodology.

PERFORMANCE INTELLIGENCE

Performance Beyond Qubit Count

Physical qubit count is the most cited but least useful single metric for comparing quantum processors. Gate fidelity, circuit depth, connectivity, and throughput matter more for real workloads.

hardware.co indexes published benchmark data across multiple metrics, normalized to source methodology where possible.

Explore Benchmarks →

Note: Not all metrics are directly comparable across architectures. See methodology for details.

Quantum Volume

A holistic measure of circuit depth, error rate, and connectivity. Higher is better. Accounts for qubit count, fidelity, and circuit transpilation simultaneously.

IBM uses QV as a primary device benchmark. Current commercial systems range from QV 32 to QV 4096+.

CLOPS

Circuit Layer Operations Per Second

Measures QPU throughput — how many quantum circuit layers can be executed per second including classical overhead. Critical for variational algorithms.

IBM introduced CLOPS in 2021. Higher CLOPS means faster iteration for VQE/QAOA experiments.

F₂Q

Gate Fidelity

Probability that a two-qubit gate operation produces the correct output state. Typically reported as average fidelity across all qubit pairs.

Superconducting: 99.0–99.5%. Trapped-ion: 99.5–99.9%. Measured via randomized benchmarking.

#AQ

Algorithmic Qubits

IonQ's metric for effective circuit capability: the largest n such that an n-qubit Quantum Fourier Transform succeeds with >50% probability.

Not directly comparable to physical qubit count. IonQ Forte: 35 AQ from 32 physical qubits.

INTELLIGENCE

QPU Intelligence

View QPU Intelligence →

Research

Trapped Ion vs Superconducting for QAOA at 100+ Qubits

Which QPU actually runs a 100-qubit QAOA job faster and cheaper: trapped ion or superconducting. Fidelity, connectivity, wall-clock, and dollar cost compared.

Samuel Edwards·Sep 29, 2026·8 min

Read →

Research

Picking a QPU for VQE Chemistry at 20 to 50 Qubits

Compare IonQ Forte, Quantinuum H2, and IBM Heron for 20 to 50 qubit VQE chemistry runs by fidelity, connectivity, circuit depth, and per-shot cost.

Eric Lamanna·Sep 22, 2026·7 min

Read →

Benchmarks

IBM Heron r2: A Performance Deep-Dive After Six Months of Cloud Access

We ran 40,000 circuits on IBM's 156-qubit Heron r2 processor. Here's what the data actually shows about two-qubit gate fidelity, crosstalk, and real workload throughput.

QPU Intelligence Desk·Jul 14, 2026·11 min

Read →

View QPU Intelligence →

FOR TECHNICAL & BUSINESS TEAMS

Evaluating Quantum Infrastructure?

hardware.co helps technical and business teams understand processor architectures, evaluate providers, compare access options, and identify hardware appropriate for experimental workloads.

From preliminary architecture selection to vendor evaluation and procurement guidance — without the vendor bias.

Request a QPU Assessment Explore Enterprise

Quantum runs next to classical.

qpuQPU vs GPUWhere each processor wins, and why most quantum workloads are hybrid.hardwareSize classical AI hardwareGPUs, memory and power for the models on the classical side.hardwareOn-prem AI hardwarecstmAI^™ systems, specified and installed by hardware.co.

NEWSLETTER

The QPU Brief

Processor launches, benchmark updates, hardware roadmaps, and quantum-computing infrastructure news. No hype.

No spam. Unsubscribe anytime. See Privacy Policy.