jiuzhang-sdk 0.1.0a30

九章光量子真机云接入软件包（jiuzhang SDK）

JiuZhang Photonic Quantum Cloud Platform Python SDK

GitHub: https://github.com/JiuZhangLiangZi/jiuzhang-sdk

jiuzhang-sdk is the Python SDK for the JiuZhang photonic quantum cloud platform. It helps Python scripts and Jupyter notebooks submit Gaussian Boson Sampling (GBS) tasks, poll task status, parse returned results, and run lightweight local analysis.

📥 Installation

Install the current pre-release:

pip install --pre jiuzhang-sdk

Install a specific version:

pip install jiuzhang-sdk==0.1.0a30

💡 Tip: If you want to experience the most complete workflow (including environment checks under Jupyter, HTML table rendering, and local sampling data analysis comparisons such as TVD, Hellinger distance, and JS divergence), it is recommended to check out the Jupyter Notebook Example (sdk_full_usage.ipynb) at the root of the repository.

🚀 Quickstart

Python:

from jiuzhang import CloudClient, GBSParams

# Configure API Key at the top of the script
api_key = "your-api-key-here"

# Initialize client
client = CloudClient(api_key=api_key)

# Build GBS task parameters directly with literal values
params = GBSParams(
    project_id="EXP-39dfbfdcab444d32",
    quantum_computer_id="PH_QC_04",
    mt=300,
    pump_energy_nj=4.6,
    task_name="GBS experiment 001",
)

result = client.run_gbs(params)
print(result.task_id)
print(result.status_name)
print(result.quantum_computer_id)

client.close()

🧩 Core Concepts

☁️ CloudClient

CloudClient is the main entry point of the SDK, used to establish HTTP communication with the remote JiuZhang Photonic Quantum Cloud Platform. It handles runtime complexity estimations, task submissions, and polling status queries.

---

⚛️ GBSParams

GBSParams is the typed task configuration object for GBS experiments.

from jiuzhang import GBSParams

params = GBSParams(
    project_id="EXP-39dfbfdcab444d32",
    quantum_computer_id="PH_QC_04",
    mt=300,
    pump_energy_nj=4.6,
    task_name="GBS experiment 001",
)

Field	Type	Required	Description
project_id	str	Yes	Project experiment_code on the cloud platform
quantum_computer_id	str	Yes	Target cloud quantum computer ID (e.g. PH_QC_04)
mt	int	Yes	Time-bin count, range 1 <= mt <= 500
pump_energy_nj	float	Yes	Pump pulse energy in nJ
squeezing_param	float | None	No	Squeezing parameter
task_name	str	No	Custom display name for the task, up to 200 characters

---

📊 GBSResult

GBSResult is the parsed object that holds GBS task states and experimental result distribution curves.

Property	Description
task_id	Unique task ID generated by the cloud platform
status_name	Normalized task status string (e.g., SUCCESS, FAILED, PENDING, RUNNING)
project_id	Associated Project ID
quantum_computer_id	Target quantum computer ID where the task ran
experimental_distribution	Experimental distribution data points (experimental inside result_map_points)
ground_truth_distribution	Reference ground-truth distribution data points (ground_truth inside result_map_points)
result_map_points	All computed probability distribution curves returned by the server
download_url	URL to download the raw sampling result data file from the cloud
raw	The original raw response JSON dictionary from the cloud API

---

🏗️ Architecture

The SDK adopts a lightweight client-only architecture, shielding users from complex quantum hardware driver orchestration:

Layer	Responsibility	Key Components
User Layer	Python scripts, Jupyter Notebooks, business applications	User Code
SDK Layer	Parameter validation, task lifecycle management, result parsing	CloudClient, GBSParams, GBSResult
Transport Layer	Request authentication, header injection, HTTP JSON transport	TokenManager, HTTP Client
Cloud Platform Layer	Quantum calculation services, classical complexity estimation	JiuZhang Photonic Quantum Cloud Platform

---

🧱 Components

• ☁️ CloudClient: Connects to the cloud service, submits tasks, queries status, and gets GBS results.
• ⚛️ GBSParams: Wraps physics and run configuration parameters for GBS GBS tasks.
• 📊 GBSResult: Parses and holds task statuses and distribution curves.
• 🔐 TokenManager: Validates API Keys and handles log masking/protection.
• 📈 jiuzhang.gbs.analysis: Local lightweight analysis toolbox (e.g., TVD, Hellinger distance, JS divergence).
• 📒 jiuzhang.jupyter: Jupyter Notebook integration helpers (HTML formatting and environment checks).

---

🌐 Cloud API Contract

The SDK interacts with the cloud service via these RESTful HTTP endpoints:

SDK method	HTTP path	Description
estimate_runtime(...) / estimate_gbs(params)	POST /estimate	Estimate classical simulation runtime
submit_task(...) / submit_gbs(params)	POST /tasks/submit	Submit a GBS GBS experiment task
get_result(task_id)	GET /tasks/{task_id}	Fetch task status and distribution curves
run_experiment(...) / run_gbs(params)	Integrated Workflow	Estimate -> Submit -> Poll -> Parse

All communication is authenticated via the X-Jiuzhang-API-Key header. The SDK parses response payloads, validates internal code status, and raises specific exceptions if code != 0.

---

📒 Jupyter Notebook Integration

The SDK provides specialized helpers to optimize Jupyter interactive workflows. For the complete Notebook example, please refer to: sdk_full_usage.ipynb.

from jiuzhang.jupyter import display_gbs_result, get_notebook_client, show_environment

# 1. View masked current environment variables
show_environment()

# 2. Get client automatically and execute a GBS task
client = get_notebook_client()
result = client.run_gbs(params)

# 3. Render a beautiful HTML table summary in the notebook cell
display_gbs_result(result)

---

🛠️ Development & Testing

Under the code/ directory:

# Format code style
ruff format src tests
# Run static checks
ruff check src tests
# Run strict static typing checks
mypy src
# Run test suite
pytest

---

📄 License

Proprietary. Copyright 2026 JiuZhang Quantum. All rights reserved.