Phase Differential Theory
Phase Differential
A Framework for Reality
Visits …

For investors

The near-term
case.

PDT’s near-term commercial programme focuses on software and model-based decision methods that can be evaluated against existing approaches. The immediate priority is prospective quantum control: identifying when an intervention is likely to help and testing whether that decision produces a measurable net benefit.

Open a conversationQuantum control dossier

The problem

When should a controller act?

Quantum devices drift and pick up noise. A controller can intervene, for example with a phase correction, but every intervention costs measurement, time and sometimes added error. The practical question is whether a controller can tell in advance when acting will help, and whether acting on that judgement gives a net improvement over the best conventional controller with the same information and resources.

The proposed approach

PERSIST-Q

Available information

Observations and records available before the decision deadline.

Intervention opportunity

A specified correction, such as phase feedforward, or no action.

Estimator

No public specification is available yet, so no estimator equation is shown here. It will be published with its definition, inputs and source version.

Comparison

The strongest implementable conventional controller and simpler baselines, under matched resources and latency.

The evidence

What exists today

  • Mathematical The record and prediction framework in the current foundations. Prediction dossier
  • Simulated None reported.
  • Experimental None reported. No benchmark report exists, and no claim is made that PDT outperforms established controllers or existing information theory.

Next milestones

Validation sequence

  1. 01

    Reproducible model and implementation

    Not yet complete

    A published specification of the estimator, controller and code.

  2. 02

    Held-out benchmark against strong matched comparators

    Not yet started

    Fixed protocol, matched resources and latency, reported deadline misses.

  3. 03

    Independent replication

    Not yet started

    Another group reproduces the benchmark from the published materials.

  4. 04

    Partner evaluation or pilot

    Not yet started

    A defined pilot on a partner's own data or device.

  5. 05

    Deployment assessment

    Not yet started

    Latency, resource costs and operational value in a realistic setting.

The next measurable step is milestone 1, followed by the held-out benchmark.

The collaboration sought

What would help most

  • Suitable, documented control or readout data.
  • Apparatus expertise for a defined qubit or memory system.
  • Help building a fair benchmark with strong comparators.
  • A defined pilot with clear success criteria.

Exploratory

Longer-term applications

These are research ideas, not products. Each lists what must be shown before a stronger commercial claim can be made.

  • Synchronisation and timing systems

    Required: A specified method that beats established synchronisation algorithms on public benchmarks.

  • Data compression

    Required: Measured gains over standard compressors on public datasets, with code released.

  • Sensing and calibration

    Required: A calibration method validated on real sensor data against established estimators.

  • Machine learning and inference

    Required: A specified method with a matched baseline showing a measured benefit.