Rasqal is a dynamic quantum hybrid runtime with one goal: to statically solve as much of the quantum program as possible before sending what remains to a quantum computer. It's heavily inspired by SAT solvers and interpreters, mixing techniques from both together to dynamically execute, analyze, and solve the IR it gets given.
Forq is an experimental, highly-scalable static analysis data structure for predicting quantum states. It sacrifices absolute precision for scalability and when states get too complex it condenses them, causing the bitstring results to become a predicted range of probabilities rather than a definite answer. When stable, will be available for testing via Rasqals experimental options.
Any papers written about them can found here, and act as a technical deep-dive into their concepts.
Install Rasqal in your favourite Python venv by running pip install rasqal. Our current testing is done with v3.10 of Python.
To run a QIR/LLVM file with a default QASM runner as a backend:
from rasqal.simulators import fetch_qasm_runner
# Create a QASM simulation backend with 20 qubits available.
runtime = fetch_qasm_runner(20)
runtime.run("path_to_ll_file")If your entry-point has arguments that map to Python primitives, you can use them like this:
from rasqal.simulators import fetch_qasm_runner
# Create a QASM simulation backend with 20 qubits available.
runtime = fetch_qasm_runner(20)
runtime.run("path_to_ll_file", ["squirrel", 42])Any results from the IR are just returned like a normal Python method as long as mapping succeeds:
from rasqal.simulators import fetch_qasm_runner
# Create a QASM simulation backend with 20 qubits available.
runtime = fetch_qasm_runner(20)
results = runtime.run("path_to_ll_file", ["squirrel", 42])For more complex scenarios check out our quick start, examples, or tests.
To start contributing you can either put up PRs with additional features and fixes, which will require you first to build the system locally, or raise bugs, improvements and feature requests via the repositories Issues tab.
This code in this repository is licensed under the BSD 3-Clause Licence. Please see LICENSE for more information.