Resonance Ventures’ Thesis on Photonics

To whom does the end of Moore’s law benefit?

Photonics applies the physics of light to computing and communications, and its core building block is the photonic integrated circuit (PIC): a microchip that combines components such as waveguides, modulators, and detectors so that photons, rather than electrons, carry and process information. The timing of photonics’ advancements matters because the engine of the semiconductor industry is running out of road. Gordon Moore’s 1965 observation is reaching physical limits on transistor size, electron leakage, and flattening clock speeds. Those bottlenecks are now widely read as signs that the foundations of growth for the semiconductor industry might be coming to an end. 

Photonics advocates argue that PICs can fill the gap with higher bandwidth and far lower energy use, and the case is strongest where AI clusters are choking on the copper wiring between chips, with rising power costs, data-centre water use, and the concentration of advanced chipmaking in Taiwan adding pressure to find alternatives. But photonics is still young by comparison. It has to reach silicon’s manufacturing efficiency at scale before it can displace legacy approaches, and much of its headline performance still comes from vendors rather than independent benchmarks.