DEEP RESEARCH · IMAGE SENSORS / SAMSUNG VS SONY

Samsung Electronics vs Sony Image Sensors: Value Chain Map for the Visual Intelligence Era

From CCD to CMOS, BSI, stacked sensors, 2-Layer Pixel, ISOCELL, ALoP, automotive, and SWIR

Written: 2025-12-28 · Semiconductor, image-sensor, and value-chain analysis · Naver Blog

Investment decisions are your responsibility. This material is research, not a recommendation to buy or sell.

0. Bottom line first

Image-sensor competition is no longer just a smartphone megapixel race. Sony protects the premium market with roughly 45-50% global share, Apple's supply chain, and BSI, stacked, and 2-Layer Pixel technology. Samsung is the No. 2 player at around 20% share, pursuing ISOCELL, 200MP sensors, ALoP, and automotive ISOCELL Auto.

Official fact: The source lists major inflection points: CCD invention at Bell Labs in 1969, Kodak's first digital camera in 1975, CMOS sales overtaking CCD around 2007, Sony's 2012 commercialization of the first stacked CMOS image sensor, and Sony's 2021 announcement of 2-Layer Transistor Pixel.

Interpretation: Sony's edge comes from pixel physics and customer lock-in, while Samsung's counterattack comes from process technology, vertical integration, and module form-factor innovation. From an investment lens, the ecosystem matters: testing, OSAT, and camera modules can benefit alongside sensor designers.

Axes of image-sensor competitionExpanding from light-to-signal chips into perception platforms

PixelBSI · stacked · 2-Layer · ISOCELL

ProcessTSMC/JASM · Samsung System LSI/Foundry

Back endDoosan Tesna · LB Semicon · KYEC · ASE

ApplicationsMobile · automotive · SWIR · on-sensor AI

As sensing becomes perception, logic, packaging, modules, and AI processing all gain value

1. Technology history: from CCD to CMOS

Source image on image-sensor history and structure

The history of image sensors is a push for efficiency and integration. CCD transfers photo-generated charge sequentially through a bucket-brigade structure and converts it to voltage at the output stage. Low noise and high image quality made CCD dominant in broadcast cameras, medical devices, and high-end digital cameras through the 1980s and 1990s.

CMOS reads signals at each pixel by integrating amplifiers and conversion circuitry. It can use standard semiconductor processes, integrate ADCs and ISPs on one chip, and consume less power. NASA JPL's APS work, semiconductor scaling, and noise-reduction technology made CMOS the mobile-era standard.

CCD

Starting point of high image quality

Willard Boyle and George Smith at Bell Labs invented CCD in 1969. Kodak's first digital camera used CCD in 1975, and Sony built the Handycam-era peak.

CMOS

Mobile standard

Low power, miniaturization, peripheral-circuit integration, and mass production are the strengths. CMOS began overtaking CCD sales around 2007 as smartphones spread.

BSI/Stacked

Sony's structural innovation

BSI lets light reach the photodiode without passing through wiring. Sony's 2012 stacked CMOS separated pixel and logic layers to raise performance.

2. Samsung vs Sony: share and customer portfolio

Source image comparing Samsung and Sony image sensors

Item	Sony	Samsung Electronics
Market position	About 45-50% global image-sensor revenue share in 2024	About 20% share, No. 2 globally
Core customers	Exclusive supplier position for Apple's iPhone series is the stable revenue source	Galaxy is the core captive demand source; Xiaomi, Vivo, Oppo, and other Chinese OEMs diversify the base
Brands/products	Exmor, LYTIA, and high-value markets such as mirrorless, broadcast, and security cameras	ISOCELL and commercialization of 100MP and 200MP sensors
Supply chain	Strengthens logic production through TSMC and the Kumamoto JASM fab	Vertically integrated System LSI design and Samsung Foundry manufacturing, with some mature nodes outsourced

3. Pixel-structure war

The two companies are competing to overcome the physical limits of pixels and maximize light capture.

Sony

2-Layer Transistor Pixel

Announced in 2021. Sony separates the photodiode layer and transistor layer onto different substrates and stacks them vertically. The source says this expands photodiode capacity, roughly doubles full-well capacity, and reduces noise. It is described as used in iPhone 15/16 series and Xperia 1 V.

Samsung

ISOCELL and 200MP

Samsung reduces crosstalk by building physical barriers between pixels, then improves light loss through ISOCELL 2.0 and ISOCELL Plus. It has deployed 0.6μm and 0.5μm-class pixels and HP1, HP2, HP3, and HP9 200MP products.

Optics

ALoP vs folded zoom

Samsung's ALoP places lenses horizontally above the prism to increase aperture and improve night quality while lowering module height.

Official fact: The source describes Samsung's pixel binning as combining 4 pixels (Tetrapixel), 9 pixels (Nonapixel), or 16 pixels (Tetra2pixel) into one larger pixel. It raises sensitivity in low light while preserving resolution in bright scenes.

4. Beyond mobile: automotive and specialty sensors

As mobile matures, Sony and Samsung are expanding into automotive, industrial sensors, and non-visible-light sensing.

Market	Sony strategy	Samsung strategy
Automotive	Creates sensor demand through the Afeela EV project with Honda and works with Mobileye and NVIDIA. It targets Tier 1 suppliers with sensors such as IMX828 and ISX038 and MIPI A-PHY integrated into the sensor.	Uses ISOCELL Auto and CornerPixel to provide over 120dB HDR and LED flicker mitigation. The source mentions references with Tesla HW 4.0/5.0, Hyundai Motor, and Mobileye.
SWIR	Launched IMX990/991 high-resolution SWIR sensors using Cu-Cu hybrid bonding to reduce pixel pitch below 5μm, targeting wafer inspection and food contaminant sorting.	The source focuses less on Samsung's own SWIR products and more on challengers such as Aeluma using compound-semiconductor integration on silicon wafers.
On-sensor AI	Intelligent Vision Sensor integrates AI processing into the sensor to reduce data transfer, protect privacy, and enable real-time inference.	Samsung expands the sensor and system-semiconductor ecosystem through automotive and mobile platforms.

5. Value chain: from sensor chip to camera module

The image-sensor industry connects IDMs, fabless companies, foundries, OSAT providers, and camera-module makers. Sony and Samsung both internalize sensor design and front-end manufacturing, but their logic, test, and module ecosystems differ.

Image-sensor value chainHow one chip becomes a finished camera module

Design/pixelSony · Samsung System LSI

Wafer/logicSony fabs · TSMC/JASM · Samsung Foundry

Package/testDoosan Tesna · LB Semicon · Nepes · KYEC · ASE

ModuleLG Innotek · Samsung Electro-Mechanics · Sunny Optical

Image quality starts in the pixel but is completed through yield, testing, OIS, and lens structure

Stage	Companies	Role in the source
OSAT/test	Doosan Tesna	Top vendor for Samsung CIS wafer testing with high share, investing in equipment amid reports of Samsung expanding Apple CIS supply
OSAT/test	LB Semicon	Moved beyond DDI back-end work into Samsung CIS and AP testing as a second vendor, investing in the Anseong plant
OSAT/test	Nepes and Nepes Ark	Strong in PMIC packaging and testing, while continuing to explore entry into CIS testing as Samsung's system-semiconductor ecosystem expands
Sony back end	KYEC, ASE Technology	Sony is expanding global OSAT cooperation as production volumes rise, despite historically preferring in-house back-end work
Camera module	LG Innotek	Strong in iPhone Pro/Max modules, sensor-shift OIS, folded zoom, and iPhone 16 tetraprism modules
Camera module	Samsung Electro-Mechanics	Core Galaxy S-series module supplier, expanding automotive camera modules for Tesla and expected to help modularize ALoP
Camera module	Sunny Optical	Chinese optical-components player expanding from price-competitive low/mid markets toward premium supply chains

6. Overall view

Even if smartphone growth slows, image sensors can keep growing because demand to replace the human eye is expanding. In mobile, larger sensors and ALoP-like form-factor innovation matter. In automotive, Level 3+ autonomy raises requirements for HDR, LFM, and sensor counts. At the edge, on-sensor AI becomes central.

My conclusion is that Sony remains the structural leader in premium sensors, while Samsung is attacking through process technology, resolution, module architecture, and automotive references. Investors should not stop at the two sensor makers; they should also track ecosystem bottleneck solvers such as Doosan Tesna, LB Semicon, Nepes, LG Innotek, and Samsung Electro-Mechanics.

Sources

Naver original