QClojure Braket

QClojure Braket contains a backend to run QClojure quantum algorithms with Amazon Braket.

It provides access to the simulators and to quantum hardware (QPUs) supported by Amazon Braket.

Disclaimers:

• experimental/alpha
  • the braket backend implementation is tested with the SV1 simulator and the IonQ Forte QPU
• Using Amazon Braket costs money even for using simulators
  • use the simulators provided by QClojure for experimentation first

⚠️ Important: AWS Braket requires an S3 bucket to store quantum task results. See S3 SETUP for configuration details.

Usage

The qclojure-braket provides an Amazon Braket backend for QClojure algorithms. To use it in your code, you need both the qclojure and the qclojure-braket dependency in your project definition.

Prerequisites

1. AWS Account and Credentials: Set up AWS credentials via environment variables, AWS CLI, or IAM roles
2. S3 Bucket: Create an S3 bucket for storing Braket task results (see S3 SETUP)

(require '[org.soulspace.qclojure.application.algorithm.bernstein-vazirani :as bv])
(require '[org.soulspace.qclojure.application.backend :as backend])
(require '[org.soulspace.qclojure.adapter.backend.braket :as braket])

; IMPORTANT: You must provide an S3 bucket for storing quantum task results.
; AWS Braket requires this for all quantum task executions.
;
; Ensure you have:
; 1. AWS credentials configured (via environment variables, CLI, or IAM roles)
; 2. Permissions to use AWS Braket
; 3. An S3 bucket with appropriate permissions in the region as the Braket device

;; Create a Braket backend with S3 configuration
;; Replace "amazon-braket-results-1234" with your actual bucket name
(def backend
  (braket/create-braket-backend {:s3-bucket "amazon-braket-results-1234"
                                 :region "us-east-1"}))

;; Select a specific device (e.g., IonQ Forte-1 QPU) for execution
(backend/select-device backend
                       "arn:aws:braket:us-east-1::device/qpu/ionq/Forte-1")

; Estimate the cost of running the Bernstein-Vazirani algorithm on the selected device
; with 100 shots. Costs will be around $0.30 per task fee + $0.08 per shot for IonQ devices,
; but may vary based on actual pricing.
(backend/estimate-cost backend
                       (bv/bernstein-vazirani-circuit [1 0 1 0])
                       {:shots 100})

; Run Bernstein-Vazirani algorithm on Amazon Braket
; with a specific secret string [1 0 1 0] and 1000 shots.
; The result will be a map with the counts of the measured results.
(bv/bernstein-vazirani-algorithm backend
                                 [1 0 1 0]
                                 {:shots 100})

Advanced Features

The enhanced Braket backend includes:

• Hardware Optimization: Automatic circuit optimization for device topology and gate constraints
• Error Mitigation: Zero-noise extrapolation, readout error correction, and symmetry verification
• Multi-Device Support: Supports Rigetti, IonQ, IQM, OQC, and QuEra devices
• Cost Estimation: Real-time pricing estimates for QPU usage

Documentation

📚 Documentation - Setup guides, tutorials, and device information

Quick Links

• 🚀 Getting Started - Overview and setup sequence
• 🔑 AWS Credentials - Authentication setup
• 🪣 S3 Setup - Required storage configuration
• 🖥️ Braket Devices - Available quantum hardware
• 📖 Tutorial - Step-by-step learning guide
• ⚙️ Terraform Setup - Infrastructure as code

See examples/enhanced_braket_usage.clj for code examples.

Copyright

© 2025 Ludger Solbach

License

Eclipse Public License 1.0 (EPL1.0)