Damavand is a quantum circuit simulator. It can run on laptops or High Performance Computing architectures, such CPU distributed architectures or multi GPU distributed architectures.

Damavand is a code that simulates quantum circuits. In order to learn more about damavand, refer to the documentation.

Development status

1. Core features:

• single node CPU ✔️
• single node GPU ✔️
• multiple nodes CPU ✔️
• single node multiple GPUs ✔️
• multiple nodes multiple GPUs ✔️

2. Documentation 🛠️ (under construction)
3. README.md 🛠️ (under construction)
4. pypi 🛠️ (under construction)

Installation

From pypi

pip3 install damavand

From sources

git clone https://github.com/MichelNowak1/damavand.git
cd damavand/
python3 setup.py install

Example

Damavand can be executed with different "apply methods". These refer to the algorithms used to update a quantum state by applying a given quantum gate. The different apply methods are:

1. brute_force brute matrix vector multiplication (requires a lot of memory).
2. shuffle smart implementation of matrix vector multiplication where the matrix is the result of a series of kronecker products (requires less memory but is still a bit slow).
3. "multithreading" runs on a single node multi CPUs.
4. "gpu" runs on a single node single GPU.
5. "distributed_cpu" runs on a multiple nodes multiple CPUs.
6. "distributed_gpu" runs on a single node multiple GPUs or on multiple nodes multiple GPUs.

The default "apply_method" is "multithreading". this can be changes by providing it as an option to the Circuit builder, as shown in the following example:

from damavand import Circuit

# initialize MPI
from mpi4py import MPI

num_qubits = 1
circuit = Circuit(num_qubits, apply_method="distributed_gpu")

circuit.add_hadamard_gate(0)

circuit.forward()
circuit.measure()

num_samples=10
circuit.sample(num_samples)

Slurm workload run:

Here is a simple slurm example that shows how to run a simulation on 2 nodes, with multithreading.

#!/bin/bash

#SBATCH --job-name=two_nodes
#SBATCH --qos=qos_cpu-dev
#SBATCH --ntasks=2
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task=40
#SBATCH --output=two_nodes.listing
#SBATCH --time=5:00

module purge

module load openmpi/3.1.4
module load cuda/10.2

srun python3 two_nodes.py

Run sbatch run.sh to lounch the script

Pennylane Integration

Damavand is linked to pennylane PennyLane, a library with many tools to approach qubit based and continuous-variable based quantum architectures.

The plugin that allows this can be found here.

With this plugin, one can use pennylane with the damavand.qubit backend.

dev = qml.device("damavand.qubit", wires=10, apply_method="gpu")

Sun rising on mount damavand