iPronics ONE

Programmable Optical Switching for AI Infrastructure

A rack-mounted OCS platform that combines optical switching with integrated control, telemetry, and APIs for deployment within AI network architectures.

Optical Performance & Scale

iPronics ONE combines optical density, fast reconfiguration, and integrated control in a compact platform for AI scale-up networks.
  • High-Density Optical Circuit Switching
  • 32+ Port Scalable Architecture
  • Non-Blocking Architecture
  • Low Optical Latency
  • Sub-Millisecond Optical Reconfiguration

Silicon Photonics, Built to Scale

Proprietary Silicon-Photonic Switch Fabric

iPronics combines a proprietary silicon-photonic switch fabric with a custom PDK optimized for low optical loss and crosstalk.

Solid-State Reliability

A solid-state silicon-photonic architecture eliminates moving mechanical switching elements. Standard silicon-photonics manufacturing processes enable scalable production and reliable operation. Per-channel optical gain control and integrated monitoring help maintain consistent optical performance across configured paths.

Manufacturing at Scale

iPronics is scaling production of the ONE Series through a dedicated manufacturing line at Fabrinet for packaging, assembly, testing, and qualification of silicon-photonic OCS systems. The manufacturing platform is designed to support volume production while meeting the quality and reliability requirements of large-scale AI infrastructure.

Active Optical Performance Control

Integrated SOA gain control and optical power monitoring actively manage path power for consistent optical performance across configured connections.

Software-Defined Control & Telemetry

gNMI/YANG and Python APIs enable software-defined provisioning, control, and operational visibility across the optical layer.

Path Provisioning

Create simplex, duplex, and bulk optical connections through software.

Open Control Interfaces

Configure, query, and automate optical connectivity through gNMI/YANG and Python APIs.

Streaming Telemetry

Monitor optical power and system state through real-time streaming telemetry.

Where iPronics ONE Fits

An optical switching layer replaces fixed physical-layer connections with programmable optical paths within and across racks, creating high-bandwidth, low-latency optical connectivity between GPUs when and where workloads need it.

01

Replace Static Connectivity

Replace fixed paths with software-configurable optical circuits.

02

Scale Within & Across Racks

Extend optical connectivity as GPU domains grow.

03

Connect Available GPU Capacity

Connect available GPU resources as capacity needs shift.

GPU Racks / Pods

Reserve capacity

Existing AI Network Fabric

Packet forwarding and general connectivity

iPronics ONE

Programmable optical switching layer

Training Workloads

Large GPU-domain connectivity

Inference Services

Latency-sensitive connectivity

Data Movement

High-volume east-west traffic

Connect Available GPU Capacity

Connect available GPU resources

Adapt physical connectivity as AI infrastructure scales.

Technical materials

For architects and engineers evaluating iPronics ONE, we provide reference materials to support planning, integration, and deployment

Product brief

Comprehensive overview of iPronics ONE capabilities and use cases

Download

Datasheet

Detailed technical specifications and mechanical drawings

Contact Us

Evaluate iPronics ONE

Map iPronics ONE to your architecture, identify priority deployment scenarios, and define the criteria that matter for evaluation.

We are disrupting data center management with a revolutionary optical networking engine
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. iPronics uses [its own and third-party] cookies to provide the services offered on this website and its functionalities, improve, analyze, and personalize the use and navigation on the website. For more information, you can review our Cookie Policy.

You can take a look to our privacy policy.