A LiDAR system does not always need a powerful optical return to understand its surroundings. Single-photon LiDAR takes the opposite approach: instead of relying on relatively strong reflected signals, it builds depth information from extremely weak optical returns, sometimes down to individual photons. That fundamental shift is turning the detector from a supporting component into the heart of the sensing architecture.

For the semiconductor industry, this matters because the performance of single-photon LiDAR increasingly depends on what happens inside the chip from SPAD pixels and microlenses to quenching circuits, time-to-digital converters and photon-processing logic.

The Real Breakthrough Is Happening at the Receiver

·         The defining component of a single-photon LiDAR system is the single-photon avalanche diode (SPAD). Operating in Geiger mode, a SPAD can produce a digital electrical response when an individual photon triggers an avalanche.

·         That makes semiconductor engineering central to the technology.

·         Researchers have already demonstrated SPAD integration using conventional CMOS processes.

·         One published design, for example, integrated a SPAD and active quenching/reset circuitry in a standard 0.18-µm CMOS process, showing how photon-level detection can be brought closer to mainstream silicon manufacturing.

·         The next challenge is no longer simply detecting a photon. It is detecting the right photon at the right time and from the right location.

512 × 512 Pixels Show Where the Architecture Is Heading

The scale of photon-counting arrays is becoming an important indicator of technological maturity.

A 2026 study demonstrated a range-gated single-photon LiDAR system using a 512 × 512 silicon SPAD array for high-speed 3D imaging. The system achieved a 43° × 43° field of view, operated with an integration time of only 0.625 milliseconds, and reached imaging rates of up to 51.3 frames per second. Researchers also reported a six-fold improvement in image acquisition speed through an improved first-photon imaging algorithm.

This is significant because the conversation is moving from isolated high-performance detector demonstrations toward large-format semiconductor imaging architectures.

For chip designers, array size now has to be considered alongside pixel pitch, fill factor, timing resolution, noise and on-chip processing.

Our most recent updated related study is available for free at this link: https://semiconductorinsight.com/report/single-photon-lidar-market-size-share-trends-market/

Time Is Becoming a Semiconductor Resource

Ø  In ordinary imaging, capturing more light can often improve the image. Single-photon LiDAR introduces another variable: time.

Ø  The arrival time of an individual photon can reveal the distance to an object. Consequently, time-to-digital converters, timing circuits and histogram-processing architectures become critical parts of the semiconductor platform.

Ø  Recent research has focused specifically on reducing the time required to acquire useful photon-counting data. At the same time, researchers are addressing pile-up effects that occur when multiple photons arrive within the detector’s recovery period. A 2025 Optica study examined pile-up correction for SPAD-based LiDAR architectures using time-to-digital conversion.

Ø  The implication is clear: future LiDAR chips will increasingly be judged not only by how many photons they detect, but by how intelligently they timestamp and interpret them.

2026 Research Is Pushing Photon Counting Into Difficult Environments

The technology is also moving beyond controlled laboratory scenes.

A 2026 study introduced a compact single-photon LiDAR prototype designed for extremely low signal-to-background conditions. The reported system weighed 12 kg or less, consumed approximately 25 W, and achieved around 18 cm depth resolution with approximately 9 mm lateral resolution. The researchers also demonstrated reconstruction under a signal-to-background ratio as low as 0.04.

Such work points toward applications where conventional sensing becomes difficult because available reflected light is weak or background illumination is strong.

That includes robotics, drones, scientific imaging, environmental monitoring and specialized industrial sensing.

Underwater Vision Adds another Semiconductor Test

Water introduces scattering that can severely complicate optical ranging.

Researchers at Heriot-Watt University have demonstrated underwater 3D imaging using a CMOS-fabricated silicon SPAD array. Their architecture uses 64 × 32 macro-pixels, with each macro-pixel containing 4 × 4 SPAD detectors, alongside multi-event time-to-digital conversion for simultaneous time-tagged photon acquisition.

The significance extends beyond underwater applications.

It demonstrates how semiconductor designers are being pushed to create architectures capable of handling multiple photon events, timing information and spatial data simultaneously.

The Next Differentiator Could Be Intelligence inside the Sensor

·         As SPAD arrays become larger, the volume of raw photon information can become difficult to process efficiently.

·         That is encouraging a shift toward smarter sensor architectures in which event filtering, dynamic resolution, histogram generation and other processing functions move closer to the detector.

·         A U.S. patent application for a LiDAR imaging system describes dynamic-resolution processing linked to a time-to-digital converter, with different memory banks selected according to incoming timing information.

·         This illustrates a broader semiconductor direction: sensor architecture and processing architecture are increasingly being designed together.

NASA Already Demonstrated Why Individual Photons Matter

The concept is not entirely new, but its capabilities are becoming increasingly relevant to today’s semiconductor ecosystem.

NASA documented airborne single-photon LiDAR technology capable of firing 32,000 pulses per second while splitting a laser into 100 beamlets. The system detected individual reflected photons and measured their round-trip timing to approximately 20 picoseconds, enabling high-resolution 3D Earth imaging while using substantially less energy than conventional imaging LiDAR approaches.

That history provides an important lesson for today’s chip developers: photon-efficient sensing is not merely a theoretical concept. The current semiconductor race is about making such capabilities smaller, faster, more integrated and easier to deploy.

What Semiconductor Companies Are Really Building Toward?

Ø  The emerging opportunity is therefore broader than a better photodetector.

Ø  The winning architecture will likely combine high-efficiency SPAD arrays, low-noise circuitry, precise timing, optical isolation, compact data paths and intelligent processing in increasingly integrated platforms.

Ø  The direction is already visible in recent patents and research covering microlens integration, metal structures, isolation techniques and SPAD-array architectures.

Ø  A 2025 patent application, for example, describes SPAD arrays incorporating microlenses, dielectric layers, metal structures and deep-groove separation to improve optical and electrical isolation between neighboring detector units.

Ø  For the semiconductor sector, that is perhaps the most important development of all.

Single-photon LiDAR is evolving from a specialized optical technique into a chip-design problem involving photons, timing, memory and computation simultaneously. As machines increasingly need to perceive weak signals in complex environments, the ability to extract useful depth information from fewer photons could become one of the most valuable capabilities in next-generation 3D sensing.