pbitlang 1.0.4

Domain-specific language for thermodynamic computing Hamiltonians

PbitLang

Domain-Specific Language for Thermodynamic Computing

PbitLang is a domain-specific language for expressing Hamiltonians (energy functions) with compile-time physics validation. It's designed for thermodynamic computing systems including P-bits, Ising machines, and quantum annealers.

Features

• Type System: Dedicated types for spins (Ising, Potts, clock, continuous)
• Lattice Geometries: Built-in support for chains, square, triangular, honeycomb, etc.
• Physics Validation: Compile-time warnings about frustration, critical temperatures, symmetry violations
• Standard Library: Common models (Ising, Heisenberg, Potts) pre-implemented
• Zero Dependencies: Core has no required dependencies

Installation

pip install pbitlang

Quick Start

import pbitlang

# Define a Hamiltonian
source = '''
hamiltonian IsingChain(n: int, J: real) -> ising on chain(n) {
    coupling: sum((i,j) in neighbors) { -J * s[i] * s[j] }
}
'''

# Compile and instantiate
model = pbitlang.compile(source)
hamiltonian = model.instantiate(n=10, J=1.0)

Example: 2D Ising Model

hamiltonian IsingSquare2D {
    lattice: square(10, 10, periodic)
    spins: ising
    
    // Ferromagnetic coupling
    energy: -sum over neighbors(i, j) {
        J * s[i] * s[j]
    }
    
    // External field
    energy: -h * sum over i { s[i] }
    
    param J: real = 1.0
    param h: real = 0.1
}

CLI Usage

# Compile a Hamiltonian file
pbitlang compile my_model.pbit

# Interactive REPL
pbitlang repl

Documentation

• Language Specification
• User Manual

Part of the Thermodynamic Computing Platform

PbitLang is part of the Thermodynamic Computing Platform, a comprehensive software stack for P-bit and thermodynamic hardware.

License

Proprietary Software - Copyright © 2024 David Johnson. All Rights Reserved.

You may install and use this software, but modification and redistribution are prohibited without written consent. See LICENSE for full terms.

Contact

• GitHub: dmjdxb
• Repository: Thermodynamic Computing Platform