Skip to content

Cudaq GPU

The CudaqGpu backend class allows you to access the CUDA-Q GPU-accelerated simulators. CUDA-Q is NVIDIA's open-source programming platform designed for hybrid quantum-classical computing. It allows integration of quantum processors (QPUs) alongside classical CPUs and GPUs, enabling them to work together to solve complex computational problems. CudaqGpu supports selecting specific simulation backends such as a state vector simulator or a tensor networks simulator. For more information, please refer to the CUDA-Q docs.

Selecting a Simulator

Three general purpose GPU simulators are available. Either chose target="nvidia" for state vector simulation, target="tensornet" for tensor network simulation, or target="tensornet-mps" for matrix-product-state simulation.

Running on Luna Servers

This simulated backend runs on Luna servers.

Initialization

Python
from luna_quantum.solve.parameters.backends.cudaq.cudaq_gpu import CudaqGpu

backend = CudaqGpu(
    target='nvidia',
    option='fp32'
)

Usage

Python
from luna_quantum.backends import CudaqGpu

backend = CudaqGpu()