How does programmable photonics work?

The iPronics SmartLight processors consists of an optical layer, an electronic layer and a software layer. The optical layer includes a programmable waveguide mesh core, high-performance building blocks and optical ports through fiber arrays.

The core of iPronics Smartlight technology is the waveguide mesh arrangement, a network of Programmable Unit Cells that route and process the light signal.

Each Programmable Unit Cell (PUC) is a 2 input- 2 output interferometric unit that we can tune using phase actuators.

The Programmable Unit Cell offers versatile functionality as a tunable coupler, combiner, variable optical attenuator, waveguide, or phase shifter, enabling the construction of intricate circuits.

This unit operates across three distinct states, allowing independent tuning of both splitting ratio and phase, empowering precise customization.

Which functions
can be performed using
programmable photonics?


Automatic functions allow to create interconnects just by identifying input port and output ports.


· discrete delay lines

· point-to-point switches

Beam splitters&
beam couplers

Automatic functions allow to create beam splitters by identifying the input port and the output port to split, as well as the splitting ratio and to create beam couplers by programmatic control of optical coupling and phase shift allows to couple all inputs to a single output, providing equalized beating paths.


· point-to-multipoint switches

· WDM (waveguide division multiplexers)


Automatic generation of filters, it is possible to generate finite impulse response filters (FIR), infinite impulse response (IIR) filters, including higher order filters and different combination of both such as CROW and SCISSORS filter configurations. It is possible to configure the central wavelength and extinction ratio of the filter.


· Dispersion compensation

· Noise removal

· Wavelength selective switches

Matrix multiplication

implementation of matrix operators on chip


· Coherent mixer

· Logic gates

· Neuromorphic computing