Smartphone

A smartphone is a mobile phone built around a general-purpose operating system, a touchscreen interface and an ecosystem of downloadable applications, with continuous access to the internet. Its defining feature is that it is a pocket-sized computer that also makes calls, so it sits at the intersection of telecommunications, semiconductor manufacturing, software platforms and consumer markets. It is the most widely used personal computing device in history, and its design choices shape how billions of people communicate, pay, navigate, work and get news.

The modern smartphone is the product of several converging developments: miniaturised integrated circuits following Moore's law, high-density lithium-ion batteries, capacitive multi-touch displays, cellular networks that moved from voice to high-speed data, and app distribution platforms. Its importance is economic as well as technical. Device shipments, component supply chains and platform fees make it one of the largest consumer technology markets, and its spread has changed patterns of access to information and financial services, while raising questions about privacy, security, environmental cost and attention.

This entry first defines the device and its boundaries, then describes its hardware, operating systems, wireless connectivity, batteries and cameras. It traces the history from early personal communicators to the post-2007 app era, examines manufacturing and market structure, and then turns to platform economics, social effects, security and privacy, environmental impact, regulation, and open questions about where the category is heading.

Definition and defining characteristics

What separates a smartphone from a feature phone

Reference glossaries usually define a smartphone by its software rather than its shape. A smartphone is a category of mobile phone that runs a complete operating system, which offers a standard interface and platform for advanced third-party apps. Handsets without that layer are grouped as feature phones (basic phones that are not smartphones). Three attributes do most of the sorting: an extensible general-purpose operating system, the ability to install and run software written by parties other than the device maker, and an input model built around a touchscreen. The first two are the substantive criteria. Touch input is the most visible feature today, but it is a consequence of the platform model rather than its cause, and early devices showed that a touch-operated phone and an app platform can come apart.

A smartphone also contains more than one operating system. The user-facing platform is usually supplemented by a second low-level proprietary real-time operating system which operates the radio and other hardware. This split is one reason the category is best described as a computer combined with radio components, rather than as a phone with added features.

Boundary cases

Tablets and phablets sit on the edge of the category. Smartphone operating system surveys treat smart gadgets such as smartphones, tablets and phablets as devices able to run the same kind of platform software, so the operating system, not the screen size, is the common denominator. Commentators also note that the lines between smartphones, mobile phones, and notebook computers have become blurred through the introduction of tablets and tablet-phone hybrids. Any cut-off between a large handset and a small tablet is therefore a convention. Wearables raise a different question: they run reduced operating systems and usually depend on a paired phone, and they are generally reported in separate categories.

The personal digital assistant is the most direct ancestor. Early hybrids combined telephone and PDA functions in one unit, and one contemporary review described the first such device as a hybrid phone and personal digital assistant. The category's origin therefore depends on whether one counts the PDA lineage or the first handset with a touchscreen, which is one reason the history of the term is contested.

How market trackers define the category

Shipment statistics depend on the categories set by research firms. IDC's Worldwide Quarterly Mobile Phone Tracker lists feature phones and smartphones as product categories and measures them in units, value and average selling price. Its company figures count branded devices only and exclude OEM sales, and its shipment counts cover new units and exclude refurbished devices. These rules change the totals. A vendor that manufactures handsets sold under other brands would appear larger or smaller depending on which rule is applied.

Thresholds also move the line within the category. IDC defines its next-gen AI smartphones by a neural processing unit delivering at least 30 TOPS on int-8 data, a hardware cut-off that can be revised as chips improve. Definitions also affect headline milestones. IDC reported that for the first time the number of smartphones shipped exceeded the 202 million feature phones shipped in the same quarter. That crossover date holds only under the tracker's particular definition and period, so comparisons across trackers or years require checking their rules.

"Smart" as a marketing term

"Smart" started as a description of programmability, and it still refers technically to a handset that combines telephony with general computing. Put simply, the word means that the device can host and run third-party software, multitask, and reach the internet. Marketing has since attached the word to features that do not depend on that architecture. The AI label is one example, as making AI the focal marketing message at recent flagship launches shows. When reading a claim about a "smart" device, the useful question is which capability is meant: general-purpose software, a specific hardware feature, or a brand label.

An early example: the IBM Simon

The IBM Simon illustrates these criteria. It was unveiled at the fall COMDEX convention in Vegas, but wouldn't be available to purchase by consumers until August 16, 1994. Sources differ on its dates: one encyclopedia entry gives 1992 as the release year, while others date public availability to 1994. It ran an MS-DOS-compatible operating system with the ability to install additional software using its PCMCIA slot, and it made calls over a cellular network. It had no app store, however; the phone came preloaded with several apps. Its interface was navigated with a stylus, and its performance was often slow. On the criteria above, it is an early partial match, which is why it is cited as a first smartphone with qualifications.

Hardware architecture

System-on-chip

Most smartphones centre on a system-on-chip (SoC), a single integrated-circuit built on one silicon semiconductor die that combines the central processor, the graphics unit, a neural processing unit (NPU), an image signal processor (ISP), a cellular modem and memory controllers. Integration is the defining trade-off of the design. Placing these blocks on one die shortens the paths data must travel, reduces power spent moving data between chips, and saves board space that a battery-limited device cannot spare. The CPU is usually a cluster of cores of differing size and efficiency, with the most demanding tasks moved to larger cores only when needed. The GPU is a massively parallel unit (GPU) for rendering and, increasingly, general computation. The NPU handles dense tensor arithmetic for machine-learning tasks such as speech recognition, scene detection and neural-network-inference; it is more energy-efficient than the GPU for these workloads, which is why it has become standard in mid-range and flagship chips. The ISP converts raw sensor output into images, performing demosaicing, noise reduction, tone mapping and, in recent designs, multi-frame fusion (image-processing). The modem implements the cellular protocol stack and radio-frequency front end, and in many designs it shares the die with the application processor.

Because a single die must serve many workloads, SoC performance is rarely judged by one number. Manufacturers and reviewers look at sustained rather than peak performance, since heat generated under load forces clock reductions, and at the efficiency of each block. Process node is also decisive: Samsung's Exynos 2600, for example, was reported as the first mobile SoC on a 2-nanometre gate-all-around process, which Samsung claimed brought modest gains in performance and efficiency over its predecessor.

Memory and storage

A smartphone keeps two kinds of memory. Working memory, usually low-power DRAM packaged close to the SoC (LPDDR), holds the data of running apps and the operating system's active pages. Its capacity determines how many applications can stay resident in the background before the system must discard them and reload them later, which is the practical basis of multitasking. Flash storage, typically NAND flash packaged as embedded storage, keeps the operating system, apps, photos and user data while the power is off. Read and write speeds of this storage affect app launch times, camera burst performance and the time needed to load large files.

In 2026 the cost of these chips has become a central pricing factor. Demand from AI data centres has absorbed much of the supply of high-bandwidth memory and server DRAM, and the same wafer capacity is also used for mobile LPDDR. TrendForce reported that LPDDR5X contract prices were expected to rise by 78–83% quarter on quarter in the second quarter of 2026. Counterpoint Research's bill-of-materials estimates, cited in trade reporting, put DRAM and NAND at 8% and 5% of a phone's cost in early 2025 and at 18% and 25% by mid-2026, while TrendForce's analysis gave a memory share of roughly 30–40%. The figures differ by method and by base year, so they should be read as a range. Gartner has projected that combined DRAM and SSD prices could roughly double by the end of 2026, and that this would raise smartphone prices by about 13% and cut shipments by 8.4% that year. IDC's forecast was harsher, with a decline of 16.7% to just over one billion units and average selling prices up by about 27.6%. Such forecasts disagree on magnitude, but they agree that entry-level models are most exposed, because memory is a larger share of their cost.

Displays and touch input

Two display technologies dominate. A liquid-crystal display (LCD) relies on a white backlight behind a layer of liquid crystals and polarising filters that modulate the light passing through each subpixel. Its backlight is always on, which limits contrast and black levels and adds thickness, but LCD panels are cheaper to manufacture and remain common in budget phones. An organic light-emitting diode (OLED) panel is self-emissive: each subpixel generates its own light, so black pixels are switched off completely. This gives high contrast, thinner modules and the option of flexible substrates, which made foldable phones possible. Its drawbacks include gradual ageing of the organic emitters and, historically, higher cost.

Variable refresh rate (VRR) is the ability to change how often the panel is updated. Early OLED phones had fixed refresh rates, typically 60 Hz. Low-temperature polycrystalline oxide (LTPO) backplanes allow a panel to move between very low rates, around 1 Hz, and high rates of 120 Hz or more. LTPO combines low-temperature polysilicon transistors, which provide high electron mobility for fast switching, with oxide transistors such as indium gallium zinc oxide (IGZO), which leak little current and hold a pixel's state for longer. Lowering the refresh rate on static content, such as a text page or an always-on clock, reduces the energy the display consumes, which is significant because the screen is among the largest power draws on a phone.

Touch input is handled by a sensing layer, usually a projected-capacitive grid of transparent electrodes. The controller measures small changes in the electric field when a conductive finger approaches, and it locates several contacts at once. Multi-touch sampling typically runs faster than the display refresh, which lets the device track fast gestures and reduce latency. Manufacturers differ in whether the sensor is laminated on top of the panel, built into the glass, or integrated within the display stack, and this choice affects thickness, optical clarity and repair cost.

Sensor suite

A smartphone carries a set of microelectromechanical and optical sensors that feed the operating system. The accelerometer measures linear acceleration along three axes, which supports screen rotation, step counting and crash detection. The gyroscope measures angular velocity and, fused with accelerometer data, gives a stable estimate of orientation, used in games, camera stabilisation and navigation. The magnetometer, usually a Hall-effect or magnetoresistive sensor, detects the Earth's magnetic field and provides a compass heading, though nearby metal and electronics can distort it. The barometer measures air pressure, from which the device estimates altitude changes; it helps with floor detection and with improving satellite fixes.

The proximity sensor, typically an infrared emitter and detector, turns off the touchscreen during calls when the phone is held against the ear. The ambient light sensor, a photodiode with filters, adjusts display brightness and colour temperature. The satellite positioning receiver tracks several global navigation satellite systems, such as GPS, Galileo, GLONASS and BeiDou, and newer models often receive two frequency bands to reduce errors caused by the ionosphere. Biometric scanners include fingerprint sensors, which may be capacitive or optical or use ultrasound beneath the display, and face recognition systems that combine an infrared projector and camera to build a depth map. Each sensor adds cost, board space and power draw, so the choice of sensor set is itself a design decision that the device's price segment largely determines.

Operating systems and software platforms

Two platforms account for nearly all smartphone software, and their design choices shape most of what follows: how apps are isolated, how they get permissions, how long devices receive fixes, and which alternatives to native apps are viable.

Android and iOS

Android is built on the Linux kernel, and much of its core code is published as open-source software through the Android Open Source Project. The product a consumer actually buys is usually a layered version of that base. Manufacturers add their own interfaces, default applications and update policies, while Google supplies proprietary services such as the Play Store and Google Play services on most devices sold outside China. The result is a family of variants that share a foundation but differ in appearance, behaviour and how long they are maintained.

iOS follows the opposite model. Apple writes the operating system, designs the chips it runs on, and controls which software reaches users, mostly through the App Store. The system derives from the same Darwin base as macOS. Tight control over hardware and software integration is the main argument for the closed approach, and the cost is that users and developers have far less room to change defaults or distribution, a point that later became the subject of regulatory dispute (see the regulation section).

Sandboxing, permissions and developer interfaces

Mobile operating systems run each app inside an application sandbox: the app executes in its own process, keeps its own storage, and cannot read another app's files or memory directly. Access to sensitive resources such as the camera, microphone, contacts or location is gated by permissions that the user grants, usually at the moment a feature is first used, and that can be revoked later. Android adopted this runtime-permission model in version 6.0 (2015), and iOS uses a comparable scheme with purpose strings shown in permission prompts.

Apps reach system functions through application programming interfaces (APIs), which expose cameras, sensors, payments and notifications in a documented and controlled form. Developers also embed software development kits (SDKs), packaged third-party code for advertising, analytics, sign-in or payments. Because an SDK runs inside its host app, it operates under the permissions the host has been granted, which is why third-party SDKs are a recurring focus of privacy analysis. Apps also declare which API level they target, so the OS can preserve older behaviour for older apps while applying newer permission rules to updated ones.

Security-update lifecycles

Smartphone software receives two kinds of updates. Security patches that close known vulnerabilities arrive far more often than major version upgrades, which tend to appear roughly once a year. How many of each a vendor promises varies widely. Google set a new benchmark with the Pixel 8 series and newer, committing to seven upgrades and seven years of security patches. Among other manufacturers, Samsung's Galaxy S24 and recent Z-series foldables carry seven years of updates, while Motorola's current Edge series offers 3 years of OS updates and 4 years of security updates. Budget and mid-range devices often receive the least. Apple ties iOS eligibility to chip generation: in iOS 26, the A13 Bionic chip or newer is required, so older models were left out, though they continue to receive security updates on iOS 18.

Support length determines usable life far more than hardware does. A phone with working hardware becomes risky to use for sensitive tasks once patches stop, so replacement is often driven by software. Analysts have estimated that consumers are now keeping their smartphones for four years, which means a three-year patch promise leaves a gap in the typical ownership period. Empirical work has also found that patches reach some devices late: a device-level study of Android security updates reported delays of up to 300 days for some of the currently supported devices. Regulation is beginning to address the gap; a Tech Advisor report states that for all new smartphone models released since 20 June 2025, manufacturers are obliged to provide software updates for at least five years after the device is taken off the market.

Cross-platform frameworks and web alternatives

Native apps are written in each platform's own languages and toolkits, which gives the fullest access to device features and the best performance. Cross-platform frameworks offer a way to reduce duplicated effort. Popular frameworks include Flutter, React Native, and Xamarin. Cross-platform frameworks compile to native or near-native code, delivering performance closer to native apps.

A progressive web app takes a different route, building the experience on web technology instead of a store-distributed package. A Progressive Web App is a web app that can add features such as installation, offline behavior, and push notifications. Offline behaviour depends on service workers, which can cache resources locally. Their limits are practical: PWAs may struggle with complex animations or computationally intensive tasks compared to native apps. Developers also face limited ability to access certain device-specific features, and performance and features can vary depending on the browser and device. These trade-offs explain why many organisations combine native, cross-platform and web delivery rather than choosing a single route.

Wireless connectivity

A smartphone is a general-purpose computer built around a stack of radio transceivers. Its cellular, local-area, short-range and satellite links are governed by different standards bodies and occupy different parts of the radio spectrum, and much of the engineering effort goes into fitting all of them into a thin metal-and-glass chassis without letting them interfere with one another.

Cellular networks have moved through a series of generations, each defined by a set of standards. Second-generation GSM introduced digital voice and text messaging in the early 1990s, but its data rates were measured in kilobits per second. Third-generation systems added packet-switched mobile internet in the early 2000s, and fourth-generation LTE raised peak rates into the hundreds of megabits per second while cutting round-trip delay to tens of milliseconds. Fifth-generation New Radio (NR), which the standards body 3GPP specifies, rests on three gains that work together: wider channels, more antennas using massive MIMO, and shorter scheduling intervals. To suit various spectrum configurations while keeping implementation complexity to a minimum, 5G features a variety of channel bandwidths, i.e., the bandwidth of the NR carrier, from 5 to 400 MHz. Latency gains came from this combination rather than from any single change, so headline figures for 5G depend heavily on which of these features a given network has actually deployed.

The frequency story is central to 5G. Standards split the new spectrum into two ranges. The range of FR1 is 450 MHz-6 GHz, which is typically called a sub-6 GHz frequency band. The FR2 is concentrated in 24.25-52.6 GHz and is generally called the millimeter wave. The trade-off is physical. Lower frequency bands can be used for providing wide-area nationwide coverage, whereas the millimeter wave or above 6 GHz bands can provide more targeted coverage in smaller areas with very high throughput. Millimetre-wave signals have shorter propagation distances and are more susceptible to blockage and weather effects due to increased atmospheric absorption and scattering. Because of this, a phone must constantly choose among bands. A 5G phone on a high-frequency link keeps monitoring signal quality and moves to another band when the link is about to degrade, a process known as adaptive beam switching. Millimetre-wave radios also rely on phased-array antenna modules that steer narrow beams toward the tower, which takes up space that designers would otherwise use for batteries or cameras.

Short-range links follow separate standards. Wi-Fi is defined by the IEEE 802.11 family, and each generation has added wider channels, denser modulation and newer frequency bands; recent phones support the 6 GHz band in addition to 2.4 and 5 GHz. Bluetooth operates in the unlicensed 2.4 GHz band and is managed by the Bluetooth Special Interest Group; its low-energy variant suits wearables, keyboards and accessories that run for months on small cells. Near-field communication (NFC) works at 13.56 MHz over a range of only a few centimetres. Phones use it for contactless payment cards, transit tickets and building access. Payment credentials are typically held in a secure element or a protected processing area, so the phone can present them to a reader without exposing the underlying data. Some devices also carry ultra-wideband radios, which allow more precise distance measurement between two devices.

Positioning uses satellite signals, but a phone seldom works from satellites alone. A standalone GNSS receiver needs to collect orbital data from the satellites before it can calculate a fix, and if the GNSS positioning starts without any known information about the user location, ephemerides, almanacs, or current time, it may take in worst case tens of minutes to acquire enough satellites in track and to calculate the position fix. Assisted GNSS (A-GNSS) closes this gap. A-GNSS works by providing the necessary data to the device via a radio network instead of the slow satellite link, essentially "warming up" the receiver for a fix. The network supplies orbital predictions and approximate position and time, and the receiver can then lock onto weak signals more quickly. The A-GPS server can make use of additional information such as round-trip timing measurements from the cellular base station to the handset in order to calculate a location where it may otherwise not be possible, for example when there are not enough GPS satellites visible. The method has limits: it doesn't work in areas with no coverage or internet link, although the receiver can still fall back to standalone operation. Regulation also shaped this design, since deployment of A-GPS in the United States is the result of the Federal Communications Commissions' (FCC) Enhanced 9-1-1 mandate.

Satellite messaging is a newer layer that uses the same basic idea in reverse: rather than a tower, the phone talks to a satellite. Apple introduced emergency satellite messaging with the iPhone 14 in 2022, and these models use Globalstar satellites to transmit emergency messages in areas where cellular service isn't available. Android devices use a different network operator, Skylo, which relays emergency and text traffic through satellites that Google Pixel 9's satellite SOS powered by Skylo uses. Because the links carry small, compressed payloads, the messages are short. If you call or text emergency services but can't connect because you're off the grid — somewhere with no cellular and Wi-Fi coverage — your iPhone tries to connect you via satellite. The main physical constraint is line of sight: to connect to a satellite, you need to be outside with a clear view of the sky and horizon. Signal acquisition is slower than a terrestrial connection, which is why the interface guides the user to hold the device steady and aimed at the sky.

All of these links are limited by spectrum licensing. Cellular spectrum has traditionally been allocated in a long-term license model where each carrier has exclusive use of the frequencies. That model significantly simplifies design by eliminating uncontrolled interference and has been the basic design assumption in virtually every cellular standard. Because bands are auctioned or assigned country by country, a phone sold in one market may lack support for the bands used by carriers elsewhere. Manufacturers therefore build a single radio front end that covers a long list of regional bands, and each model must pass certification from national regulators and carrier testing before it is sold. Antenna integration is a further constraint: the metal frame, the display and the battery all affect radiation patterns, and designers must balance the number of antennas against the space available for them, while still meeting exposure limits.

Origins and early devices (1990s to 2006)

The smartphone grew out of two lines of development that converged slowly: handheld personal digital assistants (PDAs) that gained telephone functions, and mobile phones that gained computing and data functions. Most early hybrids were sold as communicators or PDAs, and the word "smartphone" was applied to them retrospectively.

Hybrids: Simon, the Communicators and Palm

IBM's Simon Personal Communicator is usually cited as the first device with the modern combination of features. IBM showed a prototype, codenamed Angler, in late 1992, and Mitsubishi Electric manufactured the product. BellSouth sold it in the United States from 16 August 1994 at US$899 with a two-year service contract, and it was withdrawn in February 1995 after about 50,000 units. It had a touchscreen operated with a stylus, a predictive keyboard, faxing, email, and an address book, calendar and notepad. Its hardware was modest, with a 16 MHz processor, about 1 MB of RAM, and a small monochrome LCD. Sources differ on its launch date, with some histories giving 1992 (announcement) and others 1994 (sale).

Nokia's Communicator series began in 1996 with the 9000, a clamshell with a full QWERTY keyboard under the display. It used a 24 MHz Intel 386-class processor and 8 MB of memory shared among the operating system, applications and user data. Its PEN/GEOS software came from Geoworks, and it offered text-based web browsing and faxing over a digital cellular connection. The 9210 (2000) moved the line to Symbian OS, a platform descended from Psion's EPOC and partly funded by Nokia, and added a colour screen. The series ended with the E90 in 2007.

Palm's handhelds were the dominant PDA platform of the late 1990s. Palm OS 2.0 (1997) added email, and Palm OS 3.0 (1998) added infrared communication. Early Palms stored data in volatile memory, so unsaved information disappeared when the batteries ran down. Handspring's Treo, which Palm acquired in 2003, combined a Palm device with a phone, but it still relied on a stylus and added a QWERTY keyboard. Windows Mobile, descended from Windows CE and first branded for PDAs as Pocket PC 2000, took a similar stylus-driven approach across slider and clamshell handsets.

BlackBerry and the keyboard phone

Research In Motion, founded in Waterloo, Ontario, in 1984, specialised in wireless data and push email. Its first BlackBerry, the Inter@ctive Pager 950 (1999), ran on the Mobitex network and offered always-on messaging. Its QWERTY keyboard became a defining feature, and its navigation used a scroll wheel and then a trackball, with later models designed for thumb typing. Unlike Palm and Windows Mobile, BlackBerry devices required no stylus. The BlackBerry 5810 (2002) added GSM and GPRS support, and the brand's market remained largely corporate through the mid-2000s.

Why early devices stayed niche

Early smartphones were aimed mainly at enterprise users, and they were held back by bulky bodies, short battery life, slow analogue networks and immature wireless data services. The interfaces were also a barrier. The Simon's on-screen keyboard required the handset to be turned sideways for typing and could respond slowly, and stylus-based Palm and Windows Mobile devices were awkward to operate one-handed. Price was another constraint: the Simon's contract price placed it beyond most consumers.

Digital cellular foundations

The network base for these devices came from GSM, a digital cellular standard developed in Europe during the 1980s, with commercial service from 1991. GSM provided digital voice, text messaging and the SIM card, which separated subscriber identity from the handset. The Simon instead used the analogue AMPS network in the United States, which limited it to voice and slow data. Later nokia and BlackBerry devices relied on GSM data extensions such as HSCSD, GPRS and EDGE, which provided the data channel that later smartphones depended on.

The touchscreen and app platform revolution (2007 to 2012)

Apple unveiled the iPhone at Macworld in January 2007 and released it in the United States on 29 June 2007. Its main departure from contemporary handsets was the interface. The device had only a few physical buttons and relied on a large multi-touch display operated with fingers rather than a stylus. It also shipped with Safari, the web browser built on the WebKit engine, which let users see ordinary web pages instead of the reduced mobile versions that most rival phones offered. Capacitive sensing, which responds to the electrical properties of a fingertip and allows several contact points at once, made this kind of gesture input practical. Its launch also raised expectations about how a phone should respond to the user, with pinch-to-zoom and inertial scrolling becoming standard expectations within a few years.

The first iPhone had no third-party native software. Apple pointed developers toward web applications running inside Safari, and only in March 2008 did it announce a software development kit (SDK) for native apps. The App Store opened on 10 July 2008 with roughly 500 apps, about a quarter of them free, and developers paid a $99 annual fee to publish. The commercial model was the decisive element. Developers set their own prices, including zero, and kept 70 percent of sales revenue, with Apple taking 30 percent. Because a small studio could reach every iPhone owner through one storefront, without building its own billing or distribution, the model lowered barriers to entry sharply. Some early titles became commercially significant, and developers' cumulative earnings were later reported in the tens of billions of dollars. Apple added in-app purchases in 2009, initially only inside paid apps, which extended the model from one-off sales to recurring digital revenue. The effect was a feedback loop: more users attracted more developers, and more apps made the platform more attractive to users, a dynamic that illustrates network effects.

Android entered the market as an alternative with a different strategy. The Open Handset Alliance was announced in November 2007 with 34 founding members, including handset makers, chipset firms and mobile operators, and Google released the Android SDK days later. The first commercial Android phone, the HTC Dream sold as the T-Mobile G1, went on sale in October 2008. Android's platform code was distributed under a permissive open-source licence, so manufacturers could adopt it without paying a licence fee for the core system. Google's own applications and services were licensed separately, which gave the company influence without making the base system proprietary. This allowed vendors to build phones at many price points, from flagship models to low-cost devices sold in emerging markets, and it encouraged rapid adoption by manufacturers that could not afford to develop their own operating systems. The cost of this openness was fragmentation: manufacturers customised the interface and updated their devices at different speeds, which later became a recurring issue for security maintenance.

The shift away from feature phones was fast. Nokia had long been the largest handset maker by volume, and its Symbian platform was the most widely used smartphone OS of the late 2000s. It lost ground as iOS and Android grew, and by late 2010 Android had overtaken Symbian in global smartphone shipments. In 2011 Nokia chose to build its smartphone strategy around Windows Phone, a platform that had launched in late 2010 with a smaller app catalogue, and the transition was slow. BlackBerry, which had dominated corporate email on physical-keyboard devices, struggled to adapt to touch interfaces and to a market where app ecosystems mattered more than secure messaging. Its next-generation platform arrived late, and its share fell steeply. Feature phones, which relied on operator-controlled portals and Java-based applications, lost their commercial reason to exist once consumers could install software from a single store. By 2012, iOS and Android together accounted for the great majority of smartphone shipments, and the platform competition that defined the next decade had taken shape.

The period also established the economic template that later sections of this article describe: a hardware maker, a platform owner that controls distribution and takes a share of transactions, and a developer base whose depth determines the value of each device. Whether this structure amounts to an oligopoly or a form of market power that calls for regulatory response has been debated since the App Store's early years.

Batteries, power and charging

Cell chemistry and energy density

Nearly every smartphone uses a lithium-ion cell, which stores energy by shuttling lithium ions between electrodes through an electrolyte (see electrochemistry). Among commercial rechargeable chemistries, lithium-ion currently offers the highest energy density Among all commercially available secondary batteries, lithium-ion batteries offer the highest energy density. A cell's stored energy is roughly its charge capacity multiplied by its nominal voltage, $E \approx Q \times V$. A 5,000 mAh cell at about 3.85 V therefore holds roughly 19 Wh, and a 6,000 mAh cell about 23 Wh. Because capacity grows with the amount of active material that fits inside the handset, the central engineering trade-off is between capacity and thickness. Buyers resist thicker phones, so manufacturers look for more energy per cubic millimetre rather than simply adding cell volume Devices are getting more powerful, but users refuse to accept thicker phones, and we need a good battery to power all of it.

The anode is the main lever. Graphite has a theoretical capacity of about 372 mAh/g, while silicon's theoretical figure is about 4,200 mAh/g Conventional lithium polymer batteries typically use graphite for the anode, which has a theoretical specific capacity of 372 mAh/g. In contrast, silicon has a very high theoretical specific capacity of 4200 mAh/g. The obstacle is mechanical: silicon expands greatly when it absorbs lithium, which cracks the electrode and degrades the cell silicon swells significantly — up to 300% — when it absorbs lithium, cracking the electrode over cycles and degrading the battery quickly. Silicon-carbon composites respond by holding nano-silicon in a carbon matrix that buffers expansion, and silicon is kept to a minority share of the anode. Sources give gains of roughly 20–50% over graphite cells Gravimetric: Typically +20–50% (300–400+ Wh/kg in optimized cells)., and these cells have appeared in Chinese flagship phones with batteries of 6,000 to 7,500 mAh Mainstream in premium smartphones: Chinese flagships (Xiaomi, OPPO, vivo, HONOR, OnePlus, etc.) routinely use Si/C anodes for 6,000–7,500+ mAh batteries in slim designs.. Published density figures differ considerably depending on vendor and test conditions, so headline numbers need careful reading.

Power management

The phone's power budget is shared among the system-on-chip, display, radios and sensors. In CMOS logic (see cmos), dynamic switching power is approximately $P \approx C V^{2} f$, where $C$ is capacitance, $V$ supply voltage and $f$ clock frequency. Dynamic voltage and frequency scaling (DVFS) exploits this relationship by lowering frequency and voltage together when the workload is light. Because voltage enters the expression squared, the energy saved per operation is large. A governor in the operating system samples the load and moves the processor between operating points. Heterogeneous CPU designs complement DVFS: a few high-performance cores handle bursts of demanding work, while small, efficient cores handle background tasks, and idle blocks are power-gated to minimise leakage.

Heat is the constraint that limits how hard the chip can run. A phone has no fan, so heat from the processor, the 5G modem and charging circuitry spreads through graphite sheets and vapour chambers into the chassis. When on-die temperature sensors exceed set thresholds, the system reduces clock speeds or GPU load. This thermal throttling is why performance often falls during sustained gaming or long video recording. The battery management system (BMS) runs alongside, tracking cell voltage, current and temperature to prevent overcharge and over-discharge and to reduce charging current when cells become hot.

Wired and wireless charging

Wired fast charging relies on negotiation between charger and handset. The USB Power Delivery standard uses USB-C to agree on voltage and current, and many manufacturers also run proprietary protocols that exceed 100 W. Charging follows a constant-current, then constant-voltage profile: current is high while the cell is low, then tapers as it approaches full. Fast charging is therefore quick through the middle of the range and slow at the top. Heat is the central constraint. An older patent description explains that as the battery degrades rapidly due to the charging efficiency and the heat emitting reaction of the battery when a current for static current charging is high, the battery manufacturers mostly use charging of a low current (5 W).

Wireless charging uses inductive coupling, standardised as Qi and, in its newer form Qi2, with magnetic alignment. Its efficiency is lower than that of a cable, and the energy that doesn't make it to your device gets converted into heat. The effect on battery life is disputed. Some reviewers judge that all things being equal, wireless charging isn't significantly more harmful to your device or its battery than wired charging. Others note that the extra heat reduces battery health in the long term. Active cooling is a common mitigation. Samsung, Google, and other manufacturers sell wireless chargers with a built-in cooling fan to alleviate heating issues.

Ageing applies to all lithium-ion cells regardless of charging method. All lithium-ion batteries naturally lose capacity over time due to regular charge cycles. Heat accelerates this loss, so the main controllable factors are temperature during charging and the state of charge at which a phone is held. Many phones now offer charge-limiting modes that stop around 80%, and guides consistently treat prolonged time at full charge as a significant factor.

Battery replacement and longevity

Many contemporary handsets use adhesive-sealed batteries that require specialist tools, which makes replacement costly and often leads owners to replace the whole device. The European Union's Battery Regulation (Regulation (EU) 2023/1542) changes this. From that date, smartphones must have batteries that users can remove and replace themselves using simple tools, without needing professional service. The rule applies from 18 February 2027 The European Union will require all new smartphones sold from February 18, 2027, to have user-replaceable batteries. The regulation also limits proprietary tooling, since manufacturers cannot rely on proprietary or specialised tools unless these are provided free of charge with the product.

The regulation pairs replaceability with durability benchmarks, and sources differ on the exact thresholds. One summary cites a requirement that batteries keep at least 80% capacity after 800 charge cycles, while another cites at least 80% of its original capacity after 1,000 charge cycles. The regulation also contains alternative criteria, such as maintaining 83% battery capacity after 500 cycles and 80% after 1,000 cycles. Devices that meet these criteria may be exempt from end-user replacement, in which case batteries would only need to be replaceable by authorized technicians, not end users. Commentators have observed that manufacturers may respond by improving cell performance instead of changing the design, and that the actual market shift is likely to converge not on the expansion of removable batteries but on a price uptrend driven by enhanced battery performance.

The link to device life is central to the policy. Industry and regulatory projections suggest that longer use of existing handsets would reduce energy consumption, and the EU expects the combined rules to cut phone and tablet energy use by 33% by 2030, mostly by extending how long people hold onto their devices. Critics argue that removable designs make sealing against water and dust harder and add thickness, so the tension between repairability and robustness remains a live design question.

Cameras and computational imaging

A phone camera is a small optical system with a much smaller sensor than a dedicated camera, and most of its image quality comes from how those limits are managed. The sensor's light-gathering ability depends mainly on its physical area and on the size of each photosite. Phone sensors are small because the device has to fit many other components into a thin body, and their pixels are correspondingly tiny, typically around 0.8 to 1.4 μm, while full-frame camera pixels can reach roughly 8 μm. Larger pixels collect more photons, which reduces noise and improves dim-light performance. Since small pixels are noisy, manufacturers often use pixel binning: software or on-chip circuitry merges a group of adjacent photosites, commonly 2×2 or 3×3, into one "superpixel." A 200-megapixel sensor binned 4×4 therefore produces a roughly 12.5-megapixel image with much larger effective pixels. The trade-off is resolution: binning does not make a small sensor as good as a larger sensor with the same output resolution, but it substantially improves low-light sensitivity. Quad-Bayer colour filter arrays, in which each 2×2 block shares one colour, make this kind of binning practical.

Optical zoom is limited by the physical length of the lens. A conventional telephoto module stacks its elements along the optical axis, which pushes the lens out of the phone body as focal length increases. Periscope (folded) lenses solve this by placing a prism or mirror that redirects light by roughly 90 degrees, so the lens elements lie along the width or length of the phone rather than its depth. The effective focal length is therefore not bounded by the phone's thickness. Huawei's P30 Pro (2019) popularised the design with a 5× periscope module, and later flagships have used multi-prism or "tetraprism" arrangements to reach higher magnifications. Periscope modules are harder to stabilise and focus because the moving parts are located inside a long, narrow light path. Manufacturers therefore commonly place optical image stabilisation in the prism, which can shift slightly to counter hand movement, and use autofocus that moves lens groups to hold sharpness at high magnification. Digital zoom, by contrast, merely crops and interpolates the image, which is why optical and digital zoom are usually reported separately.

Computational photography compensates for the remaining physical limits. Most modern phone photos are produced by merging several frames captured in quick succession. Multi-frame high dynamic range (HDR) capture takes exposures at different brightness levels and combines them so that highlights and shadows both retain detail. Google's HDR+ approach, for example, captures a burst of short RAW frames and aligns and averages them to reduce noise. Night modes extend this principle: the device takes many short exposures, aligns them to correct for hand movement and merges them to accumulate light, producing an image that looks like a long exposure but is sharp. The hard parts are alignment when the subject or the camera moves, and handling moving objects so they do not appear as ghosts. Neural networks are increasingly used for scene segmentation, tone mapping, colour correction and detail enhancement, and some pipelines use machine-learned models to decide how strongly to denoise different regions of an image. Because so much of the final image is synthesised, critics note that results can differ from what the sensor recorded, a point that is contested in discussions of photographic authenticity.

Video capture relies on similar machinery. Electronic stabilisation uses gyroscope and accelerometer data to shift each frame and crop slightly, while optical stabilisation moves the sensor or lens elements physically. The image signal processor (ISP), a dedicated block within the system-on-chip, performs demosaicing, noise reduction, white balance, exposure control and compression, and it increasingly hands work to the neural processing unit. The combination of hardware stabilisation and software correction is what allows handheld footage to look steady.

Because phones are always carried, their cameras are widely used for documentation and reporting, including eyewitness footage of news events and recording of evidence by individuals and journalists. This use has prompted efforts to attach verifiable provenance metadata to images and to detect manipulation, since computational processing makes the boundary between recording and editing less clear. Phone cameras are also used in some regulated medical contexts, such as dermatological photography, certain ophthalmic and otoscopic attachments, and software that analyses images for clinical review. In these settings, the accuracy, calibration and regulatory status of the imaging system matter more than its consumer appeal, and the sensor and processing pipeline must be validated for the specific clinical use.

Semiconductor and manufacturing supply chain

Most smartphone brands do not make the chips inside their phones. Under the fabless model, firms such as Apple, Qualcomm and MediaTek focus on architecture, circuit layout and verification, then send finished designs to contract foundries. TSMC in Taiwan and Samsung in South Korea are the main foundries for leading-edge semiconductor logic, including application processors and modems. Only a few companies can afford the newest fabrication lines, so the design firms depend on a small set of manufacturers, and those manufacturers cluster in a few locations.

Handset assembly follows a separate geography. Mainland China supplied about 71% of global smartphone exports in 2025, while India and Vietnam each accounted for roughly one-eighth. India will need to compete for incremental manufacturing investments both with mainland China, which accounted for 70.9% of global smartphone exports in 2025 ahead of India's 12.6%, and with other emergent producers including Vietnam, which represented 12.5% of total shipments in 2025. India's growth has been concentrated in Apple's supply chain: iPhones made up more than 75 per cent (about $22 billion) of the country's smartphone exports in FY26. Export value reached $9.84 billion during April-June, compared to $7.97 billion a year earlier. Much of this expansion is policy-driven. India's share of US imports of smartphones reached 42.2% in 2025 from 13.6% in 2024, after tariffs on Chinese goods made alternatives more attractive. Production location is also reversible for individual models: Apple's first foldable phone is reportedly not going to be assembled in India, at least for the time being, which shows how new products can stay concentrated in China even as other lines move.

Memory is the most acute dependency. The world's top three suppliers of memory chips, SK Hynix, Samsung and Micron, have seen their stock prices hit all-time highs this year, and their production capacity is nearly booked out. Much of that capacity is now absorbed by high-bandwidth memory for AI servers, which earns higher margins, so phone makers compete for what remains. Chip makers are prioritizing those high-margin orders. That leaves less supply and pushes up prices for everyone else. TrendForce projected that conventional DRAM contract prices would surge 90-95% quarter-on-quarter in Q1 2026, while NAND flash prices will jump 55-60%. Counterpoint estimated that the bill of materials for phones priced under $200 rose 20% to 30% since the beginning of the year, with mid and high-end models seeing material costs rise 10% to 15%. Shortages propagate in predictable ways. Manufacturers can't get enough DRAM and NAND flash for budget devices, forcing them into an uncomfortable choice: hike prices or strip out features. IDC expects low-to-mid range Android devices to be hit hardest, because those buyers are the most price-sensitive. Display panels and specialised inputs, such as certain chemicals and minerals, face the same logic: they are produced by a small number of suppliers, so a shortage of any one of them can halt a production line regardless of how the phone was designed. Forecasts have also moved quickly. The projected 2026 price increase for average phones was revised from 3.6% made in September 2025 to 6.9%, and IDC later put it at 14%.

Trade policy now affects both location and price. In April 2025, the United States exempted smartphones and computers from its broad "reciprocal" tariffs, and the exemption covers phones classified under HTS 8517.13. Although phones are currently exempt from such tariffs, the Section 232 (electronics) review is underway. In August 2026, reporting on a proposal described the administration as exploring duties that apply directly to consumer devices manufactured with those chips, rather than only the chips themselves. Export controls operate in parallel: Washington has sought to limit Chinese access to high-bandwidth memory. Tariffs can also redirect trade quickly. Chinese smartphone exports to the US fell 72% YoY in April to just under $700 million in 2025, after the tariff escalation, which encouraged assembly to shift toward India and Vietnam.

Market structure and economics

Measuring the market

Most headline figures come from a few research firms, and their definitions shape the numbers. Shipment trackers count units that leave the maker and enter sales channels, so they record shipments do not equate to sales but are a measure of demand as they track the number of devices being sent to sales channels. Their forecasts also diverge. IDC's latest full-year projection in the sources is a 16.7% drop in 2026 to just over 1 billion units, while Counterpoint's most recent figure is a 14.3% year-over-year decline in global smartphone shipments for 2026. Gartner's February estimate was milder, at 8.4%. The spread shows how sensitive these estimates are to assumptions about memory costs and timing.

StatCounter works differently. It estimates share of web page requests by device type and operating system, so it reflects how people use devices rather than how many were sold. A device's traffic share can therefore differ sharply from its shipment share, particularly where older handsets stay in service for years.

Premium and mass-market segments

The market turned down in 2026 after a long run of growth. Global shipments declined 4.1% year-over-year to 289.7 million units in Q1 2026, which ends a 10-consecutive-quarter growth streak that began in mid-2023. In Q2, IDC recorded a 6.7% year-on-year fall to 277.5 million units.

The decline has been uneven. Apple and Samsung were the only two companies in the global top five to register a YoY growth in Q1. In Q2, Samsung also grew shipments by 8.1% year-on-year, lifting its share to 22.6%. Chinese budget-oriented brands fell back: Xiaomi dropped 26.3% after cutting low-end volume to protect margins, and OPPO shipments fell 17.5%, while vivo slid 19.4%. Cost pressure is steepest at the bottom of the price range. Counterpoint reported that BoM costs for the low end (below $200) had increased 20-30% since the beginning of the year, against 10-15% price increases in the mid and high segments. Counterpoint has also described a virtual collapse of the sub-$150 segment.

The 2026 memory cost shock

The memory shortage started in late 2025, and NAND and DRAM costs continue to rise, up over 300% YoY. For entry-level devices, memory has become the largest cost item: IDC reports it now accounts for more than 65% of the bill of materials for entry-level devices. Counterpoint found that memory costs had surpassed system-on-chip costs across smartphone price segments. The underlying cause is competition for wafer capacity, since data centers powering AI services have created massive demand for memory components, leaving phone makers at a disadvantage in bidding.

The result is fewer units at higher prices. IDC expects total market value will still grow 6.3% to $613 billion, because higher prices are now doing the heavy lifting that volume once did. Counterpoint's September outlook puts the average selling price at about $450, up 22% year over year. Analysts expect the supply crunch to persist through the second half of 2027.

Pricing, subsidies and replacement cycles

Vendors have passed costs on. In March 2026, OPPO officially announced that, effective March 16, it will raise prices for some already-released mid-to-low-end models. Longer ownership is a parallel trend: one industry report notes that replacement cycles have stretched beyond 40 months on average. Counterpoint expects the secondary market to benefit, projecting that this segment will grow 13% in 2026, as consumers priced out of new devices increasingly migrate to the secondary market in search of value.

The sources I retrieved do not document carrier subsidy practices, so the following is general description rather than sourced data. Subsidised handsets and instalment contracts lower the upfront price but tie the buyer to a carrier for the contract term, which can slow the upgrade cycle in the same way longer software support does.

Competition

Apple's position is strong in the premium tier. Counterpoint expects its global market share to climb from 20% in 2025 to 23% in 2027. IDC attributes part of the 2026 contraction to Apple's decision to delay its next base iPhone model to early 2027, which would reduce iOS shipments by more than 4%. Chinese vendors face the sharpest squeeze at the low end. Trade reporting on Huawei is limited in these sources: one headline says it reclaims top spot as China's smartphone market contracts amid rising costs, and IDC's coverage points to standout performances from iOS, HarmonyOS, and foldables. Because scale in memory purchasing and premium brand loyalty favour a few large vendors, the market is moving toward an oligopoly structure in which barriers to entry for budget makers are rising.

Platform economics and business models

App store commissions, in-app purchases and subscriptions

Smartphone platforms earn much of their revenue from software and services rather than from hardware sales. The main channel is the app store, which takes a percentage of paid transactions. Apple's standard commission on paid apps and in-app purchases of digital goods has been 30%, falling to 15% for developers in its Small Business Program (under US$1 million in annual proceeds) and for subscriptions after their first year. Rates vary by market and are being revised under regulatory pressure. In March 2026 Apple cut its standard rate on the mainland Chinese storefront from 30% to 25%. Under Apple's EU terms, a standard in-app rate of 26% is scheduled to apply from 1 October 2026, alongside separate fees for alternative payment processing and for distribution outside the App Store.

Three revenue types matter most. In-app purchases of consumable or unlockable digital content are one-off. Subscriptions generate recurring revenue, which is why the step-down after the first subscription year matters: it rewards long-retained subscribers. Physical goods and real-world services such as ride-hailing, bookings or delivery are generally exempt from the commission, so platforms do not take a cut of most e-commerce. Because commissions accrue from the installed base, platform revenue can keep growing even when device shipments fall. The gap between the two is a central reason the platforms' economics differ so sharply from the hardware business.

Carrier-linked models, installment financing and trade-ins

In the United States, the subsidised two-year contract has largely given way to device installment plans. Carriers no longer require a fixed service commitment for the phone itself; instead they attach promotions to the installment plan. Terms have lengthened: AT&T and Verizon use 36-month plans, and T-Mobile has extended its standard equipment installment plan to 36 months with a 0% annual percentage rate on some options, while warning that rising memory costs will push device prices up. Trade-in credits are typically delivered as monthly bill credits over 24 to 36 months, and they usually require the line to stay active for the full term. Apple's US carrier terms illustrate the lock-in: if the eligible service is cancelled, the credits stop and the customer owes the remaining device balance.

Shifting the device cost onto the customer benefited carriers, who previously absorbed subsidies and recovered them slowly. Carriers have also securitised installment receivables; Verizon issued a public debt offering backed by smartphone installment plans in 2016. Trade-in offers have grown in size as upgrade cycles have slowed, because carriers use them to retain existing subscribers and to win switchers. Chinese, European and other markets rely on different mixes of carrier subsidy, retail instalments and unlocked sales, so these US arrangements should not be treated as universal.

Network effects and switching costs between iOS and Android

The value of a mobile platform rises with the number of its users, because more users attract more developers, accessories and compatible services. This is a network effect. The iOS and Android ecosystems also create switching costs. A user who moves platforms may have to repurchase paid apps, lose message histories held inside a closed messaging service, reconfigure payment wallets and account settings, and give up accessories such as watches or earbuds that pair only with one operating system. Cloud-stored photos and backups add further friction. The costs are often greater than the nominal price of the handset, which helps explain why platform loyalty persists across multiple hardware generations.

These barriers are not symmetrical across users or markets. Switching is easier for users with few paid apps and no paired accessories, and harder for families or professionals who share services. Competition regulators have treated such lock-in as a barrier to entry, and the same features that keep users loyal also sustain the market power that antitrust authorities examine. The resulting disputes are discussed in the governance section below.

Wearables, cloud accounts and paid services

The smartphone increasingly acts as the hub of a wider ecosystem. Wearables typically require a companion phone for initial setup, data syncing and many features, so the phone's operating system determines which watches or health trackers a user can practically buy. Cloud accounts hold photos, documents and backups, and the storage tiers are sold as subscriptions. Music, video, fitness, payment and cloud-storage services often sit on the same account, and platform operators bundle them to raise the number of paid services per user. Each added service makes the account more costly to leave.

For platform owners, this ecosystem model turns a one-off hardware sale into a long relationship with recurring fees. It also gives a stronger argument for higher device prices, since buyers are paying for continued access to connected services as well as for the handset. Because the memory cost shock of 2026 is pushing up hardware prices, the share of revenue derived from services and subscriptions is likely to remain an important part of platform strategy.

Social and cultural impact

Access and the digital divide

Smartphones have reached a majority of people, but access is uneven. The GSMA's 2023 tally put ownership at 54% of the world's population, roughly 4.3 billion people. Smartphone ownership continues to grow, with 250 million new smartphone owners in 2024. Owning a handset does not guarantee use. The GSMA distinguishes a coverage gap, meaning people outside mobile broadband reach, from a usage gap, meaning people who live within coverage but do not go online. The usage gap is now ten times larger than the coverage gap, which makes affordability and skills more important than network build-out. About 600 million people still reach the internet through a feature phone.

The gap is also geographic and gendered. Only 25% of the population in the world's least developed countries were using mobile internet, compared with 52% across LMICs and 85% across high-income countries. Sub-Saharan Africa and South Asia hold most of the women who lack mobile internet access, and the GSMA's 2026 gender report still records a 13% gap in smartphone ownership in low- and middle-income countries. In these markets, the smartphone is often a person's first and sometimes only route online, so device price, data cost and digital literacy shape who can participate in online banking, schooling and government services.

Mobile money and platform-led finance

Mobile money is the clearest case of a phone-based service reshaping financial access. M-Pesa, Safaricom's service in Kenya, was launched on 6 March 2007, originally as an idea part-funded by the UK Department for International Development in partnership with Vodafone. It answered a practical problem: remittance services were provided by the Post Office and Western Union, but these were slow and expensive, and people regularly resorted to informal arrangements. Users could move small sums over distance without a bank account, and the service later extended to savings and credit. Kenya's formal financial inclusion rose from 26.7% in 2006 to 82.9% of the population by 2019, though the AERC case study attributes this to M-Pesa together with the agent banking network that followed it. By 2024, the service reported 60 million users and over 17 billion transactions, a figure from a commentary rather than the operator's own accounts.

Academic reviews remain cautious about causation. One survey notes that the literature has not fully established the economic impact of mobile money on financial inclusion, and it lists the take-up of mobile money; mobile money and financial inclusion; substitutability between mobile money and conventional finance as open questions.

In Asia, the analogous model is the super-app. Services such as WeChat, Alipay, Grab and Gojek bundle messaging or payments with transport, food delivery, lending and other services in one application, and their platform economics depend on the same network effects that drive mobile money adoption. Safaricom's own M-Pesa Super App, launched in 2021, follows this pattern: it is being used by 9 million customers to access a broader range of services including savings, insurance and credit.

Changes in communication, work, news and navigation

Smartphones changed everyday routines in ways that are difficult to quantify precisely. Messaging and video calls have displaced many voice calls and SMS exchanges, and social norms have shifted towards constant availability and expectations of rapid replies. Work has become more mobile, with email, collaboration tools and authentication apps extending the office into personal time and location. News consumption has moved from scheduled broadcasts and print editions towards algorithmically ordered feeds, where the distribution of stories depends on platform design. Navigation has moved from paper maps and fixed directions to continuously updated routing that uses satellite positioning and traffic data, changing how people learn places and recall routes.

Screen time, attention and adolescent mental health

Research on smartphones and young people's mental health is unresolved, and the disagreement is partly about measurement. A systematic review of 50 articles found that most have found associations between screen exposure and mental health in adolescents, yet the associations with depressive symptoms were small to very small, varying between different devices and uses. Some researchers argue that screen time is no longer a useful construct, but it still dominates research and public discourse, and they prefer measures of what young people post and view.

Study design is a central limit. Many of the existing studies on the relationship between smartphone/social media use and mental health were cross-sectional and are vulnerable to reverse causation issues. Longitudinal work offers a more mixed picture. A large-scale longitudinal study from the UK on more than 12,000 young people found that frequent social media use predicted later poor mental health and well-being. Specification-curve analyses, by contrast, have found little evidence of substantial negative associations, and their authors believe that these effects are too small to warrant policy change.

Other findings point to purpose rather than hours. One review concludes that the impairment of mental health in adolescents is closely related to the purpose of screen use and not just the exposure time, and social media use was negatively associated with mental well-being and, in girls, associated at higher risk for depression. National surveys report correlations of the same kind; for example, a South Korean study found higher odds of loneliness, depressive mood, and perceived stress among adolescents using smartphones for more than four hours per day than among those using them for under two hours. Whether these links are causal, and through which biological or social mechanisms, remains under debate, with indeterminate underlying biological mechanisms. The evidence therefore supports a description of association and context, while leaving the size of effects and their causal direction contested.

Security, privacy and data collection

A modern smartphone holds messages, photographs, payment credentials, location history, biometric templates and the authentication tokens that link a person to bank, email and cloud accounts. That concentration makes the handset a high-value target for criminals and state actors, and it also makes it a hub for data collection by app developers, advertisers and platform owners. Security and privacy on the device therefore depend on the operating system, the network, the apps and the hardware working together.

Main threat types

Malware and phishing remain the most widespread threats. Malicious apps can request permissions they do not need, and phishing messages lure users into entering credentials on fake sites. Attacks against the phone number itself are a separate category. In a SIM swap, a fraudster persuades a carrier's staff, often through social engineering, to move a victim's number onto a SIM the fraudster controls. Instead of hacking your password, fraudsters trick customer service representatives into moving your phone number to a SIM card they control. The attack works because number portability is a legitimate feature: it is normally used when a phone is lost or stolen, or a customer is switching service to a new phone. Once the number is under the attacker's control, one-time codes sent by text message can be intercepted, which is why SMS-based two-factor authentication is considered weak. Authenticator apps generate time-based codes locally on the device, so they do not rely on cellular networks and keep working even when attackers control the phone number. Newer risks include attacks on eSIM provisioning: criminals who compromise a user's mobile-carrier account can start an over-the-air profile download to an attack SIM.

Commercial spyware is a different threat, sold by private firms mainly to government clients. The best-documented example is NSO Group's Pegasus. Descriptions of it emphasise covert monitoring of a handset and extraction of its contents without any action by the owner. The most dangerous variants are zero-click: these vulnerabilities do not rely on the target doing anything at all in order for Pegasus to compromise their device. Such exploits typically target widely used messaging and calling services, which must process data from unknown senders. The company maintains that its tools are used only to investigate terrorism and crime, but Amnesty International's forensic methodology report concluded that these claims do not hold. Defences against this class of attack are limited: standard antivirus tools will not catch Pegasus, and detection usually depends on forensic analysis of device backups. Apple's Lockdown Mode, a restrictive setting for high-risk users, successfully blocked one of the three attacks examined in one Citizen Lab investigation.

Permissions, tracking and the location data trade

Mobile operating systems control app access to sensors and data through permission models. Apps must request access to location, camera, microphone, contacts and photos, and users can grant these as always, while in use, or not at all. Apps use these sensors, and their SDKs, to build profiles, so the permission model is only part of the picture. Location data is especially valuable and has been collected in bulk by data brokers, who sell or license it to advertisers, analytics firms and others. Concerns about this trade include re-identification of individuals from precise movement traces and the sale of data to parties the user never dealt with directly.

Apple's App Tracking Transparency (ATT) changed the advertising side of this economy. Introduced in iOS 14.5 in 2021, it requires that iOS apps obtain explicit user opt-in before accessing the device's advertising identifier or tracking activity across other companies' apps and websites. Without that permission, the app cannot access the identifier or share data for cross-app advertising. When the default moved from tracking-unless-disabled to opt-out, most users declined. The effects were large. The Financial Times estimated that ATT cost major platforms including Facebook, YouTube, Snapchat and Twitter a combined $9.85 billion in advertising revenue in 2021. Advertisers responded by moving to aggregated and modelled measurement. Meta responded with Aggregated Event Measurement, a privacy-preserving framework that limits the number of trackable conversion events per domain and introduces modelled reporting. Critics also point out that Apple's own advertising revenue rose after making the change, so the policy's competitive effects remain disputed. Regulators have scrutinised Apple over ATT's impact on competition, and Germany took the most consequential action.

Hardware-backed protections and regulation

Phone makers use hardware to protect the most sensitive secrets. A secure enclave or trusted execution environment is an isolated processing area that holds cryptographic keys and biometric templates, so that the main operating system and apps never see raw fingerprint or face data; they receive only a yes or no from the authentication process. Stored user data is typically encrypted with keys tied to the device passcode, and the keys are unavailable until the user unlocks the phone. Biometric authentication such as fingerprint or facial recognition is convenient, but it is usually a means of unlocking a device rather than a replacement for a passcode, and it can be overridden by coercion or by a fallback passcode.

Privacy law shapes what platforms may collect. The emergence of more robust privacy measures in everyday technology has partly been motivated by new data protection and privacy laws around the globe, particularly the GDPR in the EU and UK since May 2018. The General Data Protection Regulation requires a lawful basis for processing personal data, such as valid consent, limits collection to what is needed for a stated purpose, and gives individuals rights of access and erasure. Its effect on advertising has been uneven: consent prompts became common in apps and websites, while platform-level changes such as ATT, which operate outside the regulation's consent framework, reshaped the flow of identifiers in parallel. The result is a layered system in which legal obligations, platform policy and hardware protections each restrict data flows in different ways, and the boundaries between them continue to be contested.

Environmental and resource impact

Manufacturing versus use in life-cycle assessment

Life-cycle assessment (LCA) of a handset usually finds that production carries most of the greenhouse-gas burden when only the device itself is counted. A review of published mobile-phone studies reached this conclusion for a typical smartphone . Apple's own disclosure attributes about 81% of an iPhone 13's lifetime emissions of roughly 64 kg CO2 to production and 16% to use. Independent work points the same way: Ericsson's model of a phone with a three-year life gave a total of 57 kg CO2-equivalent, with component production above 80%, and integrated-circuit manufacturing alone about two-thirds of that share.

The answer shifts with the system boundary. In one study, a phone on its own had about 35 kg CO2-eq from manufacturing against 8.5 kg from use; once the radio network and servers were included, use rose to 47 kg, which made it the larger stage . Expanding the scope to servers and networks raised annual emissions from about 19 kg to 62 kg in one estimate . Specific disputes include:

• Boundary choices. Compared with a Nokia model, the Fairphone's manufacturing figure is markedly lower, and the review suggests this reflects assumptions about what the manufacturing stage includes rather than a real difference in production .
• Electricity mix. Because factory and use-phase electricity are assumed to differ in carbon intensity, the fossil share of the grid strongly affects the result .
• Lifetime. Use-phase emissions only overtake production after roughly three or four years of use, while average usage is often put at around two years .

Mineral inputs and due diligence

A smartphone contains small quantities of many materials. Lithium-ion cells use cobalt in their cathodes, circuit boards and solder rely on tin, capacitors use tantalum, connectors and plating use gold, and magnets in speakers and vibration motors use rare earths. Cobalt supply is concentrated in the Democratic Republic of the Congo, where artisanal mining has been linked to child labour and hazardous conditions. Tin, tantalum, tungsten and gold (often called 3TG) are the focus of conflict-mineral rules. The US Dodd-Frank Act (2010, section 1502) requires companies to disclose whether these minerals may fund armed groups. The EU regulation on conflict minerals, applicable from 2021, requires importers of 3TG to carry out supply-chain due diligence based on OECD guidance. Cobalt falls outside these 3TG regimes and is addressed mainly through industry programmes and the EU Batteries Regulation. Rare-earth supply raises separate concerns about concentrated refining capacity. Conflict-mineral rules mainly address traceability rather than volumes.

E-waste volumes and collection

The e-waste problem is larger than phones alone. In 2022 the world generated 62 billion kg of e-waste, or 7.8 kg per person, and only 22.3% of it was documented as properly collected and recycled . The generation rate is projected to reach 82 million tonnes by 2030, and the documented recycling rate could fall to about 20% if current trends continue . Rates differ sharply by region: Europe recycled 42.8% of its e-waste, while African countries documented less than 1% . Countries with binding collection targets averaged about 25%, against 0% where no legislation exists . The UN report does not break out smartphones, but its small IT and telecommunications category, which includes phones, had a recycling rate of 22% . Take-back schemes have historically been weak; one mid-2010s assessment argued that prompt return of phones would cut impacts through reuse and avoid toxic disposal .

Durability and repairability rules

The EU Ecodesign Regulation (EU) 2023/1670 has applied to new phones and tablets since 20 June 2025 . It sets a battery standard of at least 800 charge cycles while retaining 80% of initial capacity, and requires operating-system updates for at least five years from the date the last unit is sold . Key spare parts must be supplied within 5–10 working days and for at least seven years after the model stops being sold . Devices also carry a repairability score from A to E . Critics note that only five of the 15 spare-part categories must be available to end users and community repair groups, not just to professional repairers .

The link between software support and replacement is central. A phone that no longer receives security patches becomes unsafe to use for banking or messaging, so the support period effectively sets the limit of its usable life. Fraunhofer researchers expect the EU rules to lengthen the average lifespan of a mid-range smartphone from 3.0 to 4.1 years . That figure is a projection, and its realisation depends on how long manufacturers actually supply updates and parts.

Regulation, competition and governance

Europe: the Digital Markets Act and app-store steering

The most consequential regulatory action against app stores so far has come from the European Union. Under the EU Digital Markets Act (DMA), Apple is designated a gatekeeper for its App Store, and Article 5(4) requires gatekeepers to let developers steer users to offers outside the store free of charge. On 23 April 2025 the European Commission fined Apple €500 million for breaching this rule and ordered it to remove technical and commercial restrictions on steering. Apple has appealed that decision. A separate Article 6(4) investigation, which concerns distribution and interoperability, reached preliminary findings on the same day, and a final decision could carry fines of up to 10% of worldwide annual net sales. Apple responded by restructuring its EU fees, including a commission on some link-out purchases, and the Commission said it was assessing whether the new terms complied. Apple has since said it will update its EU developer terms by October 2026 to allow third-party stores and direct web distribution. The episode shows how a regulator can use fines, cease-and-desist orders and daily penalty threats to reshape fee structures, while the company contests the legal basis of each step.

United States: litigation over anti-steering rules

In the United States, the central case is Epic Games v. Apple, filed in August 2020 in the Northern District of California. In 2021 the district court rejected most of Epic's antitrust claims but found that Apple's anti-steering rules violated California's Unfair Competition Law, and it issued a permanent injunction. The Ninth Circuit affirmed that injunction in 2023. In April 2025 the same court held Apple in civil contempt, reasoning that Apple had violated the "spirit" of the injunction by charging link-out fees that were too high. The Ninth Circuit later upheld the contempt finding while allowing Apple to charge some reasonable fee tied to its intellectual property. In June 2026 the US Supreme Court agreed to hear Apple's appeal, so the question of how far a platform may charge for off-store purchases will be argued in its next term. Google faces parallel proceedings over Google Play billing and distribution, and the US Department of Justice's monopolization suits against Google's search and Android businesses form a wider backdrop, though these are not limited to app stores.

Japan and other jurisdictions

Japan's Mobile Software Competition Act took effect on 18 December 2025. The Japan Fair Trade Commission designated Apple and Google in March 2025 for their mobile operating systems, app stores and browsers, and Google for search. The law requires them to permit alternative app marketplaces, alternative payment providers and external links inside apps, and it bars them from favouring their own services. Apple and Google introduced compliance changes in iOS 26.2 and in Google Play on the same timeframe. Critics in the United States have argued that the law targets two American firms, a charge the Japanese regulator has not accepted. Other jurisdictions, including South Korea and the United Kingdom, have also moved toward ex ante rules for dominant platforms, with the UK approach relying on a designated-firm regime that is more flexible in its remedies. Across these cases the same dispute recurs: platform owners argue that commissions and review pay for security and intellectual property, while developers and regulators argue that the fees and friction suppress competition.

Standards bodies and radio certification

Regulation of the device's radio behaviour rests on a different layer. The 3rd Generation Partnership Project (3GPP) develops the specifications for cellular networks from 3G onward, and the IEEE 802.11 working group defines the Wi-Fi standards. Their specifications determine which bands, modulation schemes and protocols a phone must support, but they do not themselves grant legal permission to transmit. National radio regulators, such as the US Federal Communications Commission or the European radio equipment framework, certify that a device meets emission and interference limits before it can be sold, and they allocate the spectrum that operators use. Because each market certifies devices separately and licenses bands differently, manufacturers often ship region-specific hardware variants.

Consumer protection: warranties, repair and obsolescence

Consumer law shapes how long a phone is expected to last. Legal warranties, which in the EU and many other markets run for at least two years, sit alongside manufacturer warranties that are usually shorter. Right-to-repair laws, enacted in several US states and in the EU, require manufacturers to make parts, tools and service manuals available, and the EU's ecodesign rules for smartphones, applicable from June 2025, set minimum requirements for battery durability, spare-part availability and software updates. Debates over planned obsolescence focus on battery throttling, where Apple acknowledged in 2017 that it slowed some older iPhones to prevent unexpected shutdowns, and on the practice of ending software support, which can make a working device unsafe or incompatible. Regulators in France and the United States have examined these issues, and some cases ended in settlements rather than final rulings. Critics argue that these rules remain uneven, while manufacturers argue that complex designs make repairs harder to standardise.

Technical trends and open questions

On-device AI and neural processing units

The neural processing unit (NPU) has moved from a premium differentiator to a common block in phone chips. By one tally, over 80% of flagship phones released in 2026 include dedicated NPUs, up from roughly half just two years ago. The NPU is designed for the matrix arithmetic behind neural-network inference, much as a GPU is designed for graphics. Vendors quote its speed in trillions of operations per second (TOPS), but these numbers are hard to compare. One review gives the iPhone 17 Pro's Neural Engine 38 TOPS, while another puts Apple's A19 Pro at ~45 TOPS — more than double the A17 Pro. Qualcomm's successor is reported to deliver up to 46 percent faster AI performance compared to the previous generation.

The main design question is where a given model runs. Tasks such as photo enhancement and live translation can, according to one review, work without an internet connection, whereas advanced generative image creation, large language model queries, and some search-based AI tools still require connectivity. Local inference keeps personal data on the handset and avoids a network round trip, but it is bounded by memory capacity, memory bandwidth and the heat a phone can dissipate, which is why vendors use techniques such as model compression and quantisation. Platform strategies differ: Apple's approach leans hardest into privacy, with most features running entirely on-device, while Samsung blends on-device speed with cloud power.

Access to the NPU is also contested. One analyst argues that Qualcomm's Hexagon NPU is largely inaccessible to third-party apps in the way that Apple's Neural Engine is not, and that Android developers broadly cannot assume NPU access, so they don't build for it. A different commentary holds that the Snapdragon 8 Elite series provides open APIs that allow app developers to leverage NPU acceleration directly. The disagreement reflects the fragmentation of vendor software development kits, which determines whether third-party developers can target the hardware at all.

Foldable and rollable form factors

Foldables remain a small but growing segment. Omdia reports that global foldable smartphone shipments reached 18.2 million units in 2025, with year-on-year growth slowing to 6%; during the first half of 2026, shipments reached 5.1 million units, down 21.6% YoY. The mix is shifting: flip-type foldable smartphones declined 47% compared with 1H25, while book-type foldable smartphone shipments remained broadly flat. Omdia attributes the revival to tablet-like wide-format foldable designs from vendors such as Huawei and Samsung. Market-share estimates diverge: Smart Analytics Global puts Huawei at 48% global market share in Q2 2026, while Counterpoint's forecast gives Samsung the lead for the full year. Apple's entry is the largest pending variable. One report says Apple unveiled a $1,999 book-style foldable on September 9, 2026, though forecasts of its volumes range widely. Foldables still make up less than 2% of global smartphone shipments by volume, they make up more than 5% of total smartphone market value.

Rollable displays, which can change their active area, have been demonstrated in prototypes and small volumes but have not reached mass-market phones. Under-display sensors, such as cameras and fingerprint readers placed beneath the OLED panel, reduce the visible notch or punch-hole but depend on panel transmittance and image-processing compensation. Both remain engineering-limited rather than established categories.

Satellite connectivity and 6G

Satellite links are moving from specialist hardware into the phone's standard radio stack. The 3GPP specifications have included support for non-terrestrial networks since Release 17, which allows a handset to attach to satellites using cellular protocols. Commercial services still depend on operator agreements, limited bandwidth and clear-sky antenna geometry, so they are best understood as supplements to terrestrial coverage.

The 6G timetable is now fixed in outline. The 6G specifications will be written during Release 21, which will span from March 2027 to the end of 2028. The first 6G standards are set to be complete in early 2029, and the schedule has checkpoints at the World Radiocommunication Conference next year (WRC-27) and the Summer Olympic Games in Los Angeles in 2028. Architecture work is also significant: the 6G core network architecture, including AI-native core architecture, is scheduled to be completed in June 2028. Commercial availability is expected later still, with one paper placing deployment in the 2028–2030 window. Earlier estimates from Ericsson placed the first specifications at the end of 2028, so the schedule has slipped by a few months.

Open questions

Several questions remain unresolved. The first is whether replacement cycles will keep lengthening. Some forecasts point to continued contraction: SAG estimates that total global smartphone shipments declined 8% YoY in Q2 2026 and forecasts a 12% YoY decline for full-year 2026. Whether lower unit sales reflect longer ownership, weaker demand or cost-driven price rises is not settled by current data.

The second is whether wearables or augmented-reality glasses will displace the phone as the primary personal computing device. The evidence is still thin, and it is unclear whether these devices will supplement the phone, depend on it for connectivity and computing, or replace it for specific tasks. The third is how long software support should last. Longer security-update commitments extend usable life, but they also raise questions about hardware cost, component availability and the environmental cost of keeping older devices on the network.

References

1. Smartphone definition (Phone Scoop)phonescoop.com
2. Emulating web browser in a dedicated intermediary boximage-ppubs.uspto.gov
3. A Survey on Smartphones Security: Software Vulnerabilities, Malware, and Attacksarxiv.org
4. Smartphoneitlaw.fandom.com
5. IDC numbers shows smartphones out shipping feature phonesallaboutwindowsphone.com
6. IDC - The Future of Next-Gen AI Smartphonesidc.com
7. First Smartphone Turns 20: Fun Facts About Simontime.com
8. IBM Simonen.wikipedia.org
9. Worldwide Quarterly Mobile Phone Trackermy.idc.com
10. Worldwide Smartphone Market Grows 1.0% in Q2 2025, Despite Global Uncertainty and Weak Demand in China, according to IDCmy.idc.com
11. History of smartphonesen.wikipedia.org
12. How IBM invented the smartphone, then abandoned it - Fast Companyfastcompany.com
13. Gartner Says Surging Memory Costs Will Reduce Global PC and Smartphone Shipments in 2026gartner.com
14. Smartphone Shipments Set for Record 16.7% Drop in 2026, as the Memory Crisis Hits Full Forceidc.com
15. Smartphone Production May Drop Over 15%: 2026 Memory Surge Ignites Cost Stormtrendforce.com
16. Smartphone Chip Shipments Fell 15% and Memory Prices Are Up 300%, and You're Paying for Itandroidheadlines.com
17. Mobile DRAM Prices Surge 78% in Q2 2026 as Memory Squeeze Hits Smartphone and Electronics Production — AtlasPCBatlaspcb.com
18. Mobile DRAM prices surge 83% in Q2 2026 as AI dataexplainx.ai
19. Smartphone prices to rise in 2026 due to AI-fueled chip shortagecnbc.com
20. What is an LTPO display? Is it different from regular OLED? - Digital Trendsdigitaltrends.com
21. LTPO Technology for OLED Displays with Variable Refresh ...foro3d.com
22. OLED vs LTPO OLED vs AMOLED: Which Displa…screenruler.online
23. What Is LTPO Displaymobilepixels.us
24. Smartphone Chip Guide: How to Choose the Right One in 2026electronics.alibaba.com
25. Smartphone Technologyqualcomm.com
26. Mobile Processors: Best Smartphone SoCs 2026mobilekishop.net
27. Lever: Speculative LLM Inference on Smartphonesarxiv.org
28. How NPUs Make Your Smartphone Smarter and Faster - Inquisitive Universeinquisitiveuniverse.com
29. Here are the phone update policies from every major Android manufacturerandroidauthority.com
30. Android Phones have an Expiry Date: How to Find out When - Tech Advisortechadvisor.com
31. Didn't Get iOS 26? Your iPhone Might Be Too Old. See the Models That Support Ittech.yahoo.com
32. This article is more than 1 year oldtheregister.com
33. 16 Acar A Device Centric Analysis of Android Security Updatesftc.gov
34. Native vs Cross-Platform vs Progressive Web Apps: The Ultimate Comparison Guide - Samyotechsamyotech.com
35. Progressive Web Apps vs. Cross-Platform Mobile Apps: A Detailed Comparisonnustechnology.com
36. Progressive Web Apps vs Native Apps: Which Is Right for You?mobiloud.com
37. Native vs. Hybrid vs. Cross-Platform vs. Progressive Web Appstechbehemoths.com
38. Interference Management in 5G and Beyond Networksarxiv.org
39. What Frequencies Do 5G Millimeter Waves Useallpcb.com
40. A-GNSS positioning in wireless mesh communication systemimage-ppubs.uspto.gov
41. Assisted GNSSen.wikipedia.org
42. Coordinates : A resource on positioning, navigation and beyond » Blog Archive » Using Assisted-GNSS to locate handsets in wireless networksmycoordinates.org
43. Which Phones Have Satellite Emergency SOS in 2026 - ecoatmblog.ecoatm.com
44. Smartphone Satellite Communication: How It Works and Why It Matterstechtimes.com
45. Use Emergency SOS via satellite on your iPhone - Apple Supportsupport.apple.com
46. The Potential of Resource Sharing in 5G Millimeter-Wave Bandsarxiv.org
47. Not All 5G Is Equal: mmWave, Low-Band, and Mid-Band Explainedhowtogeek.com
48. 5G mmWave spectrum explained5gradar.com
49. Simulating continuous millimeter wave band access to a telecommunications networkimage-ppubs.uspto.gov
50. Leveraging the potential of 5G millimeter wave - Ericssonericsson.com
51. Satellite SOS and messaging FAQsverizon.com
52. Single-Shot Position: Cell-Phone Location without Ephemeris - GPS Worldgpsworld.com
53. IBM - Simonmobilephonemuseum.com
54. The History of Simon Personal Communicator: A Complete Guidehistory-computer.com
55. Nokia 9000 Communicatoren.wikipedia.org
56. Nokia 9000 / 9110 Communicatornokia.fandom.com
57. What makes a 'Nokia Communicator'? And is the idea dead?allaboutsymbian.com
58. Nokia 9000 to E90geextreme.com
59. BlackBerryen.wikipedia.org
60. The BlackBerry Curve: When Keyboards Ruled the Worldtechspot.com
61. Classic Mobile Operating Systemsallpcb.com
62. New on Yahootech.yahoo.com
63. Team 5: 1990 – 2000slideshare.net
64. The evolution of the smartphonepocketnow.com
65. IPhone (1st generation)en.wikipedia.org
66. IPhone OS 1en.wikipedia.org
67. Apple's Safari browser turns 13 years old todayappleinsider.com
68. Apple Reinvents the Phone with iPhone - Appleapple.com
69. 10 Years of App Store: A Timeline of Changes - MacStoriesmacstories.net
70. App Store launch transforms the iPhone: Today in Apple historycultofmac.com
71. The App Store turns five: A look back and forwardmacworld.com
72. Apple's App Store Has Paid Over $70 Billion To Developersdigitaltrends.com
73. HTC Dreamen.wikipedia.org
74. Open Handset Alliance facts for kidskids.kiddle.co
75. Android (operating system)ebsco.com
76. Super high capacity of silicon carbon anode over 6500 mAh g-1 for lithium batteryarxiv.org
77. This new battery tech is all the rage—so why won't Apple and Samsung use it?howtogeek.com
78. The Secret Behind Increasing Mobile Phone Battery Capacity: Silicon-Carbon Anodesneware.net
79. Why Smartphone Batteries Are Getting Bigger in 2026worknido.com
80. Silicon-Carbon(Si-C) Batteries Explained: Higher Capacity Gains vs. Degradation, Advantages, Disadvantages, Challenges, and Real-World Trade-offscyberraiden.com
81. System and method for rapidly charging battery while considering thermal stabilityimage-ppubs.uspto.gov
82. Is wireless charging bad for smartphone battery health? - Android Authorityandroidauthority.com
83. Is Wireless Charging Bad for Battery? Fact vs. Fiction - Anker UKanker.com
84. EU mandates user-replaceable batteries by 2027 - CE Reportcereport.eu
85. EU Battery Regulation: This EU rule may let you replace your phone battery yourself from 2027business-standard.com
86. Replaceable smartphone batteries in 2027: The new EU requirementecopv-eu.com
87. EU Removable Battery Rules: What Device Designers Should Knowlarge-battery.com
88. EU Rules Will Require Replaceable Smartphone Batteries by 2027reviews.org
89. EU to Mandate User-Replaceable Smartphone Batteries from Next Year, Industry May Circumvent Regulation with High-Performance Cells Instead of Removable Designseconomy.ac
90. Is Fast Charging Bad for Your Phone? A 2026 Guideelectronics.alibaba.com
91. Smartphone cameras explained: Sensors, pixel size, aperture, and moreandroidpolice.com
92. Explained: What pixel binning means for smartphone photographyastronomy.com
93. Everything you need to know about pixel-binning - why your smartphone's 200MP camera sensor might not be all it seemsamateurphotographer.com
94. Phone Camera Specs Decoded: Megapixels vs Sensor Sizebttr.reviews
95. Everything You Need To Know About Pixel Binning In Smartphone Camerasthesmartphonephotographer.com
96. I finally learned why a 12MP photo from my phone beats a 50MP onemakeuseof.com
97. What is a periscope lens and how does it work in your smartphoneandroidpolice.com
98. The Use of Periscopic Lenses in Smartphonesazooptics.com
99. What is the Difference Between a Periscope and Telephoto Camera?xda-developers.com
100. Periscope Camera Phones: How They Work and the Best Ones to Buy in 2026 - MobilityArenamobilityarena.com
101. Computational photography part I: What is computational photography?: DPReviewdpreview.com
102. iPhone Camera Processing: From Sensor Capture to Smart HDR and Neural Engine Enhancement - AppleMagazineapplemagazine.com
103. What Is Computational Photography?howtogeek.com
104. Computational Photography: How Your Camera Uses AIphotographyicon.com
105. Computational Photography :: From Selfies to Black Holes :: vas3k.comvas3k.com
106. Featured Topicsspglobal.com
107. India posts record smartphone exports of nearly $10 billion in Q1 - Social News XYZsocialnews.xyz
108. Led by iPhones, India sees highest Q1 smartphone exports by valuebusiness-standard.com
109. AI is gobbling up the world’s memory chips, sending smartphone prices to record highs, report sayscnn.com
110. Memory Crisis: Smartphones Market Is Getting 15% Shrink In 2026 - Memeburnmemeburn.com
111. Memory chip shortage threatens Asia's smartphone affordability as prices set to surgetechwireasia.com
112. IDC sees global smartphone shipments dipping in 2026 as memory costs bitefinance.yahoo.com
113. Forget the RAM crisis: New US tariff proposal could make your next smartphone even more expensiveandroidauthority.com
114. Tariff fallout: Smartphone exports from China to US plummetemarketer.com
115. India's smartphone exports touch $13.4 billion in H1FY26 up 59 pcthehansindia.com
116. Donald Trump exempts phones, computers, other electronics from reciprocal tariffsthehill.com
117. Braving Trump tariffs, smartphones ring in growth for India's exports to US - BusinessTodaybusinesstoday.in
118. Memory Chip Shortage Hits $3 Trillion Market Value: Smartphones Cost 14% More in 2026techtimes.com
119. Smartphone Prices 2026: Why Phones Are Getting Costlierunlockbase.com
120. Samsung and Apple grow in Q1 2026 as global smartphone shipments fall 4.1% YoY: IDCfonearena.com
121. IDC: Global smartphone shipments fall 6.7% in Q2 2026 amid memory crisis - Technobaboytechnobaboy.com
122. Samsung and Apple Grew in Q1 2026 as Global Smartphone Shipments Decline 4.1% amid memory constraintsbiztechreports.com
123. Chips costs to hit smartphone shipments in 2026siliconrepublic.com
124. Memory Crunch Freezes Smartphone Market: 2026 Shipments Seen Plunging 13.9% — BigGo Financefinance.biggo.com
125. Smartphone Prices to Rise 30% in 2026 Due to AI Chip Crisis << Smartphones :: Gadget Hackssmartphones.gadgethacks.com
126. Soaring Memory Chip Prices Hit Consumer Electronics, OPPO First to Officially Announce Smartphone Price Hikes — BigGo Financefinance.biggo.com
127. Smartphones face a memory cost crunch – and buyers aren't in the moodtheregister.com
128. IDC sees global smartphone shipments dipping in 2026 as memory costs bite By Reutersinvesting.com
129. Worldwide Smartphone Market to Decline 13.9% in 2026 as Memory Crisis and US-Iran War Constrain Growthidc.com
130. Counterpoint Research Revises Down 2026 Smartphone Shipment Forecasts Amid Memory Shortage - EE Times Asiaeetasia.com
131. Smartphone Prices 2026: Why Your Phone Could Cost Moremaintenanc.me
132. Adjustments to the China storefront of the App Store on iOS and iPadOS - Latest News - Apple Developerdeveloper.apple.com
133. Apple overhauls its EU App Store fees, loosens rules for alternative app storestechcrunch.com
134. In App Purchase Fees 2026: Apple 30%, EU 26%, Google 15%appsvaluemobile.com
135. App Store Fees 2026: Apple 30% vs 15%, Google Play Ratesweekonelabs.com
136. Everything to Know About Carrier Financing Dealswhistleout.com
137. Phone Trade-In Deals: Upgrade to the Latest iPhone 18 Pro for Freewhistleout.com
138. Financing and Credit - Appleapple.com
139. T-Mobile Stretches Phone Financing to 36 Months as Device Subsidies Fade — BigGo Financefinance.biggo.com
140. iphone backed bondsfortune.com
141. Smartphone Shipments Expected to Fall 15% in 2026 as Memory Prices Push Up Phone Costs - gHacks Tech Newsghacks.net
142. Trends in Mobile Internet Connectivity - M4D - SOMICgsma.com
143. Smartphone owners are now the global majority, New GSMA report reveals - Newsroomgsma.com
144. Global mobile internet growth stallscomputerweekly.com
145. M-Pesavodafone.com
146. MOBILE MONEY, FINANCIAL INCLUSION AND DEVELOPMENT: A REVIEW WITH REFERENCE TO AFRICAN EXPERIENCE - Ahmad - 2020 - Journal of Economic Surveys - Wiley Online Libraryonlinelibrary.wiley.com
147. A Digital Financial Services Revolution in Kenya: The M-Pesa Case Studyaercafrica.org
148. What is M-Pesa? A Revolutionary Change in Africa's ...conduitpay.com
149. The associations between screen time and mental health in adolescents: a systematic reviewlink.springer.com
150. Beyond Screen Time: The Different Longitudinal Relations between Adolescents’ Smartphone Use Content and Their Mental Health - PMCpmc.ncbi.nlm.nih.gov
151. Smartphone screen time reduction improves mental health: a randomized controlled triallink.springer.com
152. Beyond Screen Time: The Different Longitudinal Relations between Adolescents’ Smartphone Use Content and Their Mental Healthncbi.nlm.nih.gov
153. Association Between Smartphone Screen Time and Mental Health Among Korean Adolescents: A National Representative Study - PMCpmc.ncbi.nlm.nih.gov
154. Adolescent screen time, anxiety/depression, and alcohol/e-cigarette use: evidence from the ABCD study - PMCpmc.ncbi.nlm.nih.gov
155. The Mobile Gender Gap Report 2026gsma.com
156. Addressing digital inclusion and the usage gapgsma.com
157. 4 Ways to Prevent SIM Swapping Attacksfingerprint.com
158. SIM swap attacken.wikipedia.org
159. What Is SIM Swapping? Attack, Definition, Preventionproofpoint.com
160. SIM-swap fraud: Scam prevention guidespecopssoft.com
161. How Does Pegasus Work?occrp.org
162. Forensic Methodology Report: How to catch NSO Group’s Pegasus - Amnesty Internationalamnesty.org
163. In-Depth Analysis of Pegasus Spyware and How To Detect It on Your Mobile Devicesgroup-ib.com
164. Israeli firm NSO’s spyware again hacking iPhones: Reportaljazeera.com
165. Apple App Tracking Transparency (ATT) in 2026: Key Business Impactspandectes.io
166. What is App Tracking Transparency?pirg.org
167. How Apple's App Tracking Transparency Impacts Measurement (And What To Do)ruleranalytics.com
168. arxiv.orgarxiv.org
169. What is SIM Swapping Fraud and How to Prevent Ittrendmicro.com
170. Method, apparatus, and computer program product for secure two-factor authenticationimage-ppubs.uspto.gov
171. What commercial spyware is, and what different types there arekaspersky.com
172. Pegasus (spyware)britannica.com
173. Redefining scope: the true environmental impact of smartphones?link.springer.com
174. Integrating Environmental and Social Life Cycle Assessment: Asking the Right Question - Suckling - 2017 - Journal of Industrial Ecology - Wiley Online Libraryonlinelibrary.wiley.com
175. Device Life Cycle Assessment → Termproduct.sustainability-directory.com
176. Life cycle environmental impacts of a smartphone - Ericssonericsson.com
177. Life Cycle Assessment (LCA) of a Smartphoneplanete-energies.com
178. The global e-waste monitor 2024 - Stichting OPENstichting-open.org
179. Global e-Waste Monitor 2024: Electronic Waste Rising Five Times Faster than Documented E-waste Recyclingunitar.org
180. The Global E-waste Monitor 2024itu.int
181. The Global E-Waste Monitor 2024 Report - Embedded Artistryembeddedartistry.com
182. Papers by James Succklingsurrey.academia.edu
183. New EU rules for durable, energy-efficient and repairable smartphones and tablets start applying - Internal Market, Industry, Entrepreneurship and SMEssingle-market-economy.ec.europa.eu
184. Europe Rewrites the Rules on Smartphones What the EU's New Repair and Battery Mandates Mean for Consumers and Manufacturerslawstreet.co
185. Repair labels, spare parts and longer support for smartphones and tablets as of June 2025 - but we need more repairable designs - Right to Repair Europerepair.eu
186. New standards for durable electronics: The EU energy label for mobile devices - Fraunhofer IZMizm.fraunhofer.de
187. Redefining Scope: The True Environmental Impact of Smartphones?search.issuelab.org
188. (PDF) Life Cycle Assessment of a Smartphoneresearchgate.net
189. Apple Inc. - Form 10-Q - FY2026sec.gov
190. Apple reveals complex system of App Store fees to avoid EU fine of 500 million eurocnbc.com
191. Apple to comply with Digital Markets Act by October 2026newsbytesapp.com
192. Apple has a month to comply with Digital Markets Act, or get fined againappleinsider.com
193. Apple faces potential Digital Markets Act fine over App Store rulespinsentmasons.com
194. Epic Games v. Appleen.wikipedia.org
195. Supreme Court Will Hear Apple's Appeal in Epic Games App Store Fight - MacRumorsmacrumors.com
196. Apple Asks Supreme Court to Review App Store Contempt Ruling - MacRumorsmacrumors.com
197. Apple Inc. v. Epic Games, Inc.oyez.org
198. Epic Games’ Ninth Circuit Win Affirming Civil Contempt Findingcravath.com
199. Japan’s “Smartphone Law” Goes Live: Apple and Google Lose Their Defaultsrealgaijin.substack.com
200. Japan’s Mobile Software Competition Act Is Now in Force — What Comes Next?law.kuleuven.be
201. Google and Apple face Japan's toughest mobile platform rules yetppc.land
202. Japan App Store Gets Alternative Marketplaces, Third-Party Payments and More - MacRumorsmacrumors.com
203. Japan cracks open the App Store and Google Play to rival payment providers and marketplacespocketgamer.biz
204. Apple, Google to allow third-party app stores under new Japanese lawaa.com.tr
205. Appcharge Blogappcharge.com
206. Best AI Phones in 2026: Top Smartphones for AI Featuresfast.io
207. On-Device AI: Revolutionizing Smartphones in 2026specser.com
208. On-Device AI Explained: How Snapdragon 8 Gen 5’s NPU Will Change Smartphones in 2026 - Gizmochinagizmochina.com
209. On-Device AI in 2026: Why TOPS Don't Tell the Whole Story - Next Waves Insightnextwavesinsight.com
210. Best Phones for AI Professionals 2026o-mega.ai
211. Omdia: Foldable smartphone shipments will double to 36 million units by 2028omdia.tech.informa.com
212. SAG: Global Foldable Smartphone Shipments Fell 2% YoY in Q2 2026; Huawei Led with 48% Share While Samsung Grew 25% - Smart Analytics Globalsmartanalyticsglobal.com
213. Foldable Phone Market Share Statistics: iPhone Duo vs Z Fold 8 - Memeburnmemeburn.com
214. 6G standardization milestones and RAN decisions - Ericssonericsson.com
215. It's officiallightreading.com
216. 3GPP approves timelines for Release 21 which will specify 6G RAN, Core and 5G Advancedtechblog.comsoc.org
217. Operator-Controlled 6G: From Connectivity Infrastructure to Guaranteed Digital Servicesarxiv.org
218. Best AI Smartphones of 2026: Top On-Device AI Phones Rankeddocam.io
219. Foldable smartphone shipments could pass 100 milliontbreak.com
220. Foldable Phone Statistics 2026: Shipments, Market Share & Forecasts - Axis Intelligenceaxis-intelligence.com
221. 6G standardization timeline and principles - Ericssonericsson.com