kanad-compute 0.2.4

Turn your computer into a Kanad quantum chemistry compute server

kanad-compute

Turn your computer into a quantum chemistry compute server for Kanad.

Run VQE calculations, molecular simulations, and quantum analyses on your own hardware — then connect to the Kanad platform for visualization, collaboration, and reporting.

Quick Start

# Install
pip install kanad-compute

# Initialize (creates config + API key)
kanad-compute init

# Start the server
kanad-compute start

Then paste your API key into Kanad > Profile > Backend Credentials > Kanad Compute.

What It Does

kanad-compute runs a local FastAPI server that executes quantum chemistry calculations using the Kanad framework. When you select "Kanad Compute" as a backend in the Kanad web app, your calculations run on YOUR machine instead of cloud services.

Supported solvers: PhysicsVQE, HardwareVQE, HybridSubspaceVQE, SQD, KrylovSQD, VQE, VarQITE, qEOM, EfficientVQE, ExcitedStates

Supported backends: Statevector (local), Qiskit Aer (CPU/GPU), IBM Quantum (with your credentials), IonQ (with your credentials)

Requirements

• Python 3.11+
• 8GB+ RAM recommended
• Kanad library installed (pip install kanad or install from source)

Installation

From GitHub (recommended)

# Clone the repo
git clone https://github.com/mk0dz/kanad-compute.git
cd kanad-compute

# Install in development mode
pip install -e .

# With GPU acceleration
pip install -e ".[gpu]"

# With IBM Quantum hardware
pip install -e ".[ibm]"

# With IonQ
pip install -e ".[ionq]"

# Everything
pip install -e ".[all]"

Note: You also need the Kanad framework installed. If you don't have it:

git clone https://github.com/mk0dz/kanad.git
cd kanad && pip install -e .

From PyPI (coming soon)

pip install kanad-compute

CLI Commands

kanad-compute init

Initialize configuration. Creates ~/.kanad-compute/config.json with a unique node ID and API key.

kanad-compute init --port 7440 --max-qubits 20 --gpu

Options:

• --port — Server port (default: 7440)
• --max-qubits — Maximum qubits to accept (default: 20)
• --gpu / --no-gpu — Enable GPU acceleration
• --ibm-token — IBM Quantum API token
• --ionq-key — IonQ API key

kanad-compute start

Start the compute server.

kanad-compute start --host 0.0.0.0 --port 7440

kanad-compute status

Check server status and system info.

kanad-compute status

kanad-compute key

Display your API key (for pasting into Kanad app).

kanad-compute key

kanad-compute configure

Update configuration without reinitializing.

kanad-compute configure --ibm-token YOUR_TOKEN --max-qubits 25

Connecting to Kanad

1. Run kanad-compute start
2. Copy your API key: kanad-compute key
3. Go to kanad.xyz > Profile > Backend Credentials
4. Under "Kanad Compute", paste the API key and set the server URL (http://localhost:7440)
5. Click "Test" to verify the connection
6. Select "Kanad Compute" as your backend when running experiments

API Endpoints

Method	Endpoint	Description
GET	/health	Server health check (no auth)
GET	/info	System info and capabilities
POST	/jobs	Submit a calculation job
GET	/jobs/{id}	Get job status and results
POST	/jobs/{id}/cancel	Cancel a running job
GET	/jobs	List recent jobs

All endpoints except /health require Bearer token authentication.

Architecture

Your Machine
+-- kanad-compute server (FastAPI, port 7440)
|   +-- /health, /info
|   +-- /jobs (submit, poll, cancel)
|   +-- Thread Pool Executor
|       +-- Kanad Solvers (PhysicsVQE, HardwareVQE, ...)
|       +-- Kanad Backends (statevector, aer, ibm, ionq)
+-- Config: ~/.kanad-compute/config.json

Kanad Web App (kanad.xyz)
+-- Profile > Backend Credentials > Kanad Compute
+-- Schrodinger Lab > Select "Kanad Compute" backend
+-- Jobs proxied to your machine via API key auth

Configuration

Config is stored at ~/.kanad-compute/config.json:

{
  "node_id": "uuid",
  "api_key": "your-api-key",
  "port": 7440,
  "max_qubits": 20,
  "max_workers": 2,
  "gpu_enabled": false,
  "ibm_api_token": null,
  "ionq_api_key": null
}

License

Apache 2.0 — see LICENSE.

Links

• Kanad Platform
• Kanad Framework
• DeepRealm Labs