Skip to content

Latest commit

 

History

History
163 lines (113 loc) · 6.67 KB

README.md

File metadata and controls

163 lines (113 loc) · 6.67 KB

copy Duke ARTIQ Extensions (DAX) for self learning

DAX is a library that extends the capabilities of ARTIQ while maintaining a vanilla ARTIQ experience. This project was initially created as a framework to develop modular control software for ARTIQ-based quantum control systems. As the project evolved, additional components and utilities were added to the repository. Users can implement modular control software for their ARTIQ projects using the DAX framework or use other components and utilities provided by DAX in existing projects.

Currently, DAX consists of the following main components:

Component Development
DAX.experiment (system) Stable
DAX.sim Stable
DAX.scan Stable
DAX.servo Beta
DAX.scheduler Alpha
DAX.program Alpha

Projects related to DAX:

DAX.experiment (system)

The DAX.experiment (system) module is a framework to develop modular control software. Using this framework, control software can be organized into modules and services with the help of a central searchable registry. Additionally, code portability can be achieved using interfaces and clients.

DAX.sim

The DAX.sim module is a functional simulator for ARTIQ hardware that allows users to run experiment code without the need for a physical core device. DAX.sim can help users test, debug, and verify the functionality of their code using existing testing environments. DAX.sim does not depend on other components of DAX and can be used by any ARTIQ project.

DAX.scan

The DAX.scan module contains lightweight scanning tools that can be used for n-dimensional scanning type experiments. The scanning tool provides an experiment template for yielding a single point of data and automates the process of scanning over one or multiple parameters. DAX.scan is not dependent on other components of DAX and can be used by any ARTIQ project.

DAX.servo

The DAX.servo module contains tools for servo control flow that can be used for closed-loop experiments with feedback. The servo class serves as a template for experiments in which a single iteration of the experiment has to be described. The servo class automates the process of looping, data handling, and exit routines. DAX.servo is not dependent on other components of DAX and can be used by any ARTIQ project.

DAX.scheduler

The DAX.scheduler module contains a toolkit for automatic scheduling of experiments. The scheduling toolkit includes classes to define jobs that can submit experiments. The provided scheduler takes a job set and schedules jobs accordingly based on their interval, dependencies, and the chosen scheduling policy. DAX.scheduler is not dependent on other components of DAX and can be used by any ARTIQ project.

DAX.program

The DAX.program module contains base classes that provide an operation-level API for DAX systems. Using these base classes, users can write operation-level programs with an API independent of the underlying system. The programs themselves should be targeted towards a specific DAX system to work within the constraints of the system and yield the best run-time performance. A DAX program is designed like a regular ARTIQ experiment and works with the same timing and execution principles. The execution model of DAX.program follows the accelerator model.

DAX.util

A collection of utilities that might be handy for any ARTIQ project.

Resources

Installation

Usage

Users can import the DAX system components and the ARTIQ experiment environment at once using the following import statement:

from dax.experiment import *

Users that would like to use DAX.sim can do so by annotating their device DB:

from dax.sim import enable_dax_sim

device_db = enable_dax_sim(enable=True, ddb={
    # Your regular device db...
})

The scanning tools in DAX.scan can be imported using the following import statement:

from dax.scan import *

The servo tools in DAX.servo can be imported using the following import statement:

from dax.servo import *

The scheduling tools in DAX.scheduler can be imported using the following import statement:

from dax.scheduler import *

The base classes and utilities for DAX.program can be imported using the following import statement:

from dax.program import *

Versioning

The major version number of DAX matches the version of the targeted ARTIQ release.

Testing

Use pytest (installed separately) to run the DAX unit tests.

pytest

Main contributors

  • Leon Riesebos (formerly Duke University)
  • Brad Bondurant (Duke University)

Publications

Acknowledgements

The development of DAX is primarily funded by EPiQC, an NSF Expeditions in Computing (1832377), and the NSF STAQ project (1818914). The work is also partially funded by the IARPA LogiQ program (W911NF-16-1-0082) and the DOE ASCR Testbed QSCOUT.

More information about these projects can be found at: