---
title: "iPronics’ $125M Series B: Optical Circuit Switching for AI Clusters"
description: "Maverick Silicon and Light Street Capital co-led iPronics’ $125M Series B — with NVIDIA participating — to scale programmable silicon-photonics switches for AI data-center networks."
date: 2026-09-04T00:00:00.000Z
tags: ["2026-vc-news", "startup-funding", "venture-capital", "ai-infrastructure", "semiconductors", "europe-vc"]
source: https://venturecapitaltracker.com/2026-ipronics-125m-series-b-optical-ai
---

# iPronics’ $125M Series B: Optical Circuit Switching for AI Clusters

> Maverick Silicon and Light Street Capital co-led iPronics’ $125M Series B — with NVIDIA participating — to scale programmable silicon-photonics switches for AI data-center networks.

**Maverick Silicon** and **Light Street Capital** co-led **iPronics’ $125 million** Series B announced **September 2, 2026**, with **NVIDIA** joining Triatomic Capital, Bosch Ventures, Catalight Capital, the European Innovation Council Fund, and existing backers (Unite.AI, Reuters). Total funding reaches **$177 million**.

**Spine:** adding GPUs only helps if bits can move between them — optical circuit switching is becoming a first-class AI infra category, not a photonics science project.

## Key facts

| Field | Detail |
| --- | --- |
| Company | [iPronics](/startup/ipronics) (Valencia, Spain) |
| Round | **$125M** Series B |
| Co-leads | Maverick Silicon, Light Street Capital |
| Strategic | NVIDIA |
| Total | **$177M** |
| Product | ONE Series optical circuit switches (32→256 ports roadmap) |
| Mfg | Dedicated Fabrinet line (announced Mar 2026) |
| Customers | First **10** hyperscaler / AI-cluster OEM engagements (company, Mar 2026) |

## Who uses the product — and for what job

**Users:** hyperscalers and AI-cluster OEMs designing scale-up/scale-out GPU fabrics where copper/electronics switching burns power, latency, and complexity.

**Job:** establish **programmable optical paths** between accelerators that can reconfigure as workloads change — iPronics cites sub-300-microsecond reconfiguration and sub-30-nanosecond latency on ONE-32 materials — instead of static cabling and ever-larger electronic packet fabrics.

## Why now

- Cluster sizes push networking from “plumbing” to a limiter of useful FLOPs.
- March 2026 Fabrinet manufacturing line was the commercialization gate; Series B finances volume and U.S. presence (Santa Clara).
- NVIDIA’s check is the signal: interconnect is part of the GPU system roadmap.

## Why Maverick / Light Street / NVIDIA — portfolio fit

Maverick Silicon is already active in AI silicon-adjacent hardware (see also Lyte). Light Street brings public/private markets hardware discipline. NVIDIA invests where networking unlocks more GPU attach.

**Likely founder rationale:** raise commercialization capital with silicon-specialist leads plus the ecosystem vendor every AI cluster buyer already standardizes on.

Maverick Silicon and Light Street are **not** in our `/fund/` directory.

| Dimension | Fit |
| --- | --- |
| Stage | Series B commercialization |
| Thesis | Photonic OCS for AI clusters |
| Geography | EU deep-tech → U.S. go-to-market |
| Risk | Ethernet/proprietary AI fabrics; manufacturing yield; long sales cycles |

## Competitive map

| Player | Lane |
| --- | --- |
| Classic Ethernet AI NICs/switches | Packet-centric scale-out |
| Other OCS / silicon photonics startups | Direct category peers |
| Vendor-integrated GPU fabrics | Vertical lock-in alternative |

## Takeaway

iPronics’ **$125M** Series B is a bet that **programmable light** becomes as strategic as the accelerators it connects. The unresolved proof: volume deployments at hyperscale economics — not lab demos — against incumbent networking roadmaps.

### Sources

1. Unite.AI: https://www.unite.ai/ipronics-raises-125m-series-b-to-scale-optical-networking-for-ai-data-centers/
2. Communications Today (Reuters wrap): https://www.communicationstoday.co.in/ipronics-raises-125m-with-nvidia-backing-for-data-centre-chip/

**By:** [Venture Capital Tracker](https://venturecapitaltracker.com/editorial-policy)

**Editorial note:** AI tools assisted with research, structure, or drafting. Venture Capital Tracker retains human editorial responsibility for factual accuracy, relevance, and source quality before publication.