Between InGaAs and Si: SWIR Sensor Supply-Chain Risks and the Practical Alternative of NIR CMOS
language
eng
date
Jun 20, 2026
slug
between-ingaas-and-si-swir-nir-cmos
author
status
Public
tags
Camera
Machine Vision
Embedded Vision
Semiconductor
SWIR
NIR
InGaAs
CMOS
Image Sensor
summary
As supply-chain risks grow around InP and InGaAs-based materials, not every SWIR application necessarily requires an InGaAs sensor. Below 1,100 nm, silicon-based BSI CMOS can be a practical alternative in terms of cost and supply stability. This article looks at the technical boundary between InGaAs and silicon.
type
Post
thumbnail
category
Camera
updatedAt
Jun 20, 2026 01:44 AM
The front line of the U.S.โChina conflict has widened, but one of its fiercest battlegrounds remains the semiconductor industry. One reason the United States may find it difficult to impose even stronger restrictions on China is that such measures could trigger Chinese export controls on raw materials and specialty materials. That would not be only a U.S.โChina issue; it could directly affect semiconductor production worldwide.
Recently, as China has tightened export controls on compound-semiconductor materials such as InP (indium phosphide), supply-chain risks have also increased for InGaAs-based SWIR sensors. Sonyโs SenSWIR sensor family, which has played a major role in expanding the use of InGaAs SWIR sensors, is not completely insulated from this trend. Considering that some key demand for SWIR sensors overlaps with military and security applications, material supply-chain issues can become even more sensitive.
Against this backdrop, Basler is promoting cameras based on OMNIVISIONโs OG05C BSI CMOS sensor as a cost-effective alternative to Sony SenSWIR sensors for imaging below 1,100 nm. Strictly speaking, this is closer to the NIR band than to SWIR, but if the required wavelength range stays below 1,100 nm, it can still be a realistic choice. With the same resolution and pixel size as the popular IMX992, a lower price, and speeds around 115 fps, not every application may need to insist on an InGaAs sensor.
The Boundary Between InGaAs and Silicon
The reason InGaAs sensors are powerful is clear. They provide differentiated information in the 900โ1,700 nm range, especially around 1,300 nm and 1,550 nm, where silicon CMOS has difficulty detecting light. In semiconductor inspection, moisture and foreign-material inspection, security and surveillance, and some military applications, InGaAs-based SWIR sensors are used almost as a standard solution.
However, not every application needs the full true-SWIR range. If sufficient information can be obtained in the NIR region around 850 nm, 940 nm, 1,050 nm, or 1,100 nm, the story changes. In this range, high-sensitivity BSI CMOS sensors can become an option in terms of price, supply stability, frame rate, and compatibility with existing camera platforms. Sonyโs SenSWIR sensors are attractive because they use Cu-Cu wafer bonding to achieve sensitivity across the visible and SWIR ranges. But if the dedicated 1,100โ1,500 nm SWIR region is not truly required, the OG05C may be usable as a lower-cost image sensor that covers a narrower but relevant range compared with SenSWIR.
If the characteristics of the inspection target are sufficiently visible below 1,100 nm, it would of course be more accurate to call this wavelength range NIR rather than SWIR. In that case, it is worth evaluating NIR CMOS-based options before adopting an expensive SWIR camera.
Therefore, rather than treating silicon as a complete replacement for InGaAs, the choice depends on whether the application needs visible + NIR or visible + SWIR. If the application depends on absorption characteristics around 1,300 nm or 1,550 nm, requires full silicon transmission, or directly uses material information available only in longer-wavelength SWIR, then a silicon-based CMOS sensor is physically close to impossible as a substitute. Ultimately, the key question is not โDo I need a SWIR sensor?โ but โDo I really need wavelength information beyond 1,100 nm?โ
As supply-chain risks around InP and InGaAs-based materials increase, sensor selection criteria will inevitably change as well. Since these components are closely related to military and security domains, it is likely to become increasingly difficult to source InGaAs image sensors under the current geopolitical environment. That makes it essential to understand exactly what the application truly needs. The gray area between InGaAs and silicon is likely to become increasingly important.
Related posts
Overview of LSSTCam: Structure and Performance of the World's Largest Camera
Jan 26, 2026
The LSSTCam is an observational camera with 3.2 gigapixels, 189 sensors, and processes tens of GB of data per second. This article briefly and clearly summarizes the key specifications and structure of the LSSTCam, explaining why it is called the 'world's largest camera.'
Cerebras WSE-3: Why Turn an Entire Wafer into a Single AI Chip?
Aug 12, 2026
Cerebras challenges the memory-bandwidth bottleneck in AI inference with WSE-3, a wafer-scale processor integrating 4 trillion transistors and 44 GB of on-chip SRAM. This post looks at the benefits of wafer-scale design, its defect-tolerant architecture, partnerships with OpenAI and AWS, and the competitive landscape with NVIDIA and Groq.
Computer Vision in the World Cup: From Goal-Line Decisions to Semi-Automated Offside
Jul 6, 2026
World Cup officiating technology is moving beyond simple video review and turning stadiums into real-time data spaces. This post explains how goal-line technology, semi-automated offside, connected-ball data, and optical tracking systems help improve both fairness and the flow of the game.