devqubit-cudaq 0.1.12

devqubit adapter for CUDA-Q

devqubit-cudaq

NVIDIA CUDA-Q adapter for devqubit — automatic kernel capture, target snapshots, and result logging for CUDA-Q sample and observe workflows.

[!IMPORTANT] This is an internal adapter package. Install via pip install "devqubit[cudaq]" and use the devqubit public API.

Installation

pip install "devqubit[cudaq]"

Usage

Sampling

import cudaq
from devqubit import track

@cudaq.kernel
def bell():
    q = cudaq.qvector(2)
    h(q[0])
    x.ctrl(q[0], q[1])
    mz(q)

with track(project="cudaq-experiment") as run:
    executor = run.wrap(cudaq)
    result = executor.sample(bell, shots_count=1000)

Observe (Expectation Values)

from cudaq import spin

hamiltonian = spin.z(0)

with track(project="cudaq-vqe") as run:
    executor = run.wrap(cudaq)
    result = executor.observe(bell, hamiltonian)
    print(result.expectation())

What's Captured

Artifact	Kind	Role
Kernel JSON	cudaq.kernel.json	program
Kernel diagram	cudaq.kernel.diagram	program
MLIR (Quake)	cudaq.kernel.mlir	program
QIR	cudaq.kernel.qir	program
Counts / expectation	result.cudaq.output.json	result
Execution envelope	devqubit.envelope.json	envelope

Documentation

See the Adapters guide for details on observe workflows, performance tuning, and spin operator capture.

License

Apache 2.0