How Lightspring Aims to Make Photonic Chip Packaging Scalable

Lightspring is addressing a problem that comes after a photonic circuit is made: connecting it reliably to fibres, lasers and other components. The Besançon-based startup uses 3D lithography to form optical interfaces directly on chips, with its September 22, 2026 financing intended to move that process from research toward industrial production. 4617

Why packaging is the target

Photonic chips need optical connections to function in a larger system. Lightspring’s target is the packaging step that makes those connections, rather than the fabrication of the underlying circuits. Reporting puts packaging at roughly half the cost of a photonic chip, making it an economically significant obstacle even when the chip itself can be manufactured. That cost share is a reported estimate, not a demonstrated saving from Lightspring’s process. 46

The question matters for systems that place optical connections near computing hardware, including proposed co-packaged-optics designs. Lightspring’s approach could simplify integration if it proves repeatable across the connections such systems require. That is a potential application, not evidence that the company has qualified a high-volume process. 1719

How Lightspring’s process works

Lightspring uses two-photon polymerisation, a form of 3D lithography, to fabricate structures such as microlenses, waveguides and beam-shaping elements directly on a chip. Instead of treating every optical connection primarily as an assembly and alignment task, the company aims to make it a programmable fabrication step. Its own description says connections can be written directly in a single process pass. 118

The company also describes process recipes, component designs and software for optical interconnects. A process design kit, or PDK, could help designers reuse coupling structures rather than develop each interface from scratch. Lightspring identifies fibre-to-chip and laser-to-chip connections among its applications. 11921

Lightspring is targeting losses below 0.5 dB per interface, according to reporting on the funding announcement. That is a target, not an independently verified result across qualified production. The supplied evidence likewise does not establish that the process works with every material or piece of equipment, or that it eliminates alignment work in every use case. 1718

What the funding pays for—and what it does not prove

The €3.1 million seed round was led by Atlantic and OVNI Capital, with Concept Ventures, Ixcore, Plug and Play Ventures and industry business angels participating. A funding listing separately records a €350,000 grant. Those two amounts add up to €3.45 million; other public funding is reported to bring total financing above €3.5 million. One secondary account identifies the grant as an i-Lab award, but the supplied excerpts do not independently establish the precise amount of the additional public funding. 481417

The money supports the transition from a research-derived technique to an industrial offering. Lightspring’s work has roots at the FEMTO-ST Institute, and reporting describes plans to expand its team to 12 and scale fabrication. Its design tools and process recipes could make customer co-design and repeat production easier to pursue; neither funding nor a roadmap, however, constitutes manufacturing qualification. 361920

The next test is practical: whether Lightspring can demonstrate optical loss, reliability, throughput, yield and cost per interface under customer-relevant production conditions. Until those results are available, its strongest claim is a promising way to tackle the packaging bottleneck—not that the bottleneck has already been solved. 1718