Skip to main content
Version: next

Examples

Prerequisites

These examples require Linux with an NVIDIA GPU and CUDA.

  • Install QDP support: pip install qumat[qdp]
  • Optional deps used below: torch, numpy

Example 1: Basic Encode + DLPack to PyTorch​

import torch
import qumat.qdp as qdp

engine = qdp.QdpEngine(device_id=0)
qtensor = engine.encode(
    [0.5, 0.5, 0.5, 0.5],
    num_qubits=2,
    encoding_method="amplitude",
)

# DLPack capsule is single-consume; convert once.
tensor = torch.from_dlpack(qtensor)
assert tensor.is_cuda
assert tensor.shape == (1, 4)  # [batch_size, 2^num_qubits]