Skip to content

Results are limited to the current section: Qpu & emulators

Product news

Emulators


Emulators reproduce neutral-atom dynamics in classical software. Use them to develop and debug sequences, study how an algorithm scales, and validate noise behaviour before committing QPU time.

All emulators are open-source and can either be run locally or accessed through Pasqal’s Cloud platform and third party cloud providers.

Emulators provide a simulation of quantum sequence execution, returning results in the same format as QPU hardware. All emulators which run on Pasqal’s cloud infrastructure are accessed using the same Cloud SDK as the QPU.

TypeQubitsCostExecution
EMU_FREE
exact20FreeRemote
EMU_MPS
approximate80PaidRemote & Local
EMU_SV
exact25–27PaidRemote & Local

Availability

Not all emulator backends are available in all environments. Check the table below for remote availability per access channel.

Pasqal CloudGoogle CloudAzure QuantumOVHCloudScaleway
EMU_FREE
EMU_MPS
EMU_SV
EMU_TN

Emulators provide a simulation of quantum sequence execution, returning results in the same format as QPU hardware. All emulators run on Pasqal’s cloud infrastructure (except local backends) are accessed using the same Cloud SDK as the QPU. A Pasqal Cloud account is required for all cloud-hosted emulators.

EMU_FREEEMU_FREE · State vector
Local
Remote
The entry-level emulator. Best used as a starting point for validating workflows and testing small-to-medium sequences before scaling to GPU-backed backends. Available at no cost to all cloud users.
Best forFirst experiments, teaching and quick noiseless sanity checks before moving to a paid backend.
Class name (Remote)RemoteEMUFreeBackend
Class name (Local)EMUFreeBackend
MAX QUBITS≤ 12 remote, ≤ 25 local
METHODState vector
EXECUTIONRemote, Local
PRICINGFree (remote), Free (local)
EMU-MPSEMU_MPS · Matrix Product State
Local
Remote
A GPU-accelerated tensor-network emulator using a Matrix Product State representation. By bounding the bond dimension it emulates far larger systems than a state vector allows, trading a controllable approximation for scale.
Best forLarge registers (max 80 qubits) with moderate entanglement, for scaling experiments toward QPU size.
Class name (Remote)RemoteMPSBackend
Class name (Local)MPSBackend
MAX QUBITS80
METHODMatrix Product State
EXECUTIONLocal · Remote
PRICINGPaid (remote), Free (local)
EMU-SVEMU_SV · State vector (exact)
Local
Remote
A GPU-accelerated exact state-vector emulator. EMU-SV keeps the full quantum state, so it returns numerically exact expectation values and bitstring distributions.
Best forHigh-precision algorithm development and research where exact state vector amplitudes are required.
Device type nameRemoteSVBackend
Class name (Local)SVBackend
MAX QUBITS≤ 25
METHODState vector (exact)
EXECUTIONLocal · Remote
PRICINGPaid (remote), Free (local)
EMU-TNEMU_TN · Tensor network
Remote
The original cloud tensor-network emulator. Superseded by EMU-MPS for most workloads but kept available for compatibility with existing pipelines.
Best forLegacy large-register emulation on Pasqal Cloud for sequences with bounded entanglement.
Class name (Remote)RemoteTNBackend
Class name (Local)TNBackend
MAX QUBITS≤ 100
METHODTensor network
EXECUTIONRemote
PRICINGPaid

Pricing

Like our QPUs, cloud emulators are available through the major hyperscalers on pay-as-you-go pricing models, so you pay only for the compute you use. Alternatively, you can buy credits directly through our sales team for use on Pasqal Cloud. For current rates, packages and to talk to sales, see our Offers & Pricing page.

View offers & pricing

Last updated: