Next-Generation Wi-Fi and Mobile Networks

When data traffic increases, not only do data centers need upgrades, but the standardized network protocols that connect devices and transmit data may also need enhancements.

Your current devices probably connect via Wi-Fi or cellular data. The "G" in 4G/5G stands for "generation," with upgrades occurring roughly every decade, from 2G for voice and short message service (SMS) in the 1990s, to 3G for media in the 2000s, to high-speed 4G in the 2010s, and 5G in the 2020s. Similarly, Wi-Fi standards are constantly evolving.

When the variety and volume of wireless data increases, wider bandwidth (larger "traffic lanes") and new frequency bands (new "routes") may be needed. This is like expanding roads or building new highways as traffic volume grows. Current Wi-Fi and 4G speeds may seem sufficient, but the surge in data may eventually drive the adoption of higher standards. As 5G and future 6G networks expand, speeds could reach 20 to 100 times that of 4G. Technologies like non-terrestrial networks (NTN), which use satellites as base stations, could further enhance network coverage.

Similarly, the transition from Wi-Fi 6/6E to Wi-Fi 7 enables faster data transmission through wider channels.

In particular, thanks to Multi-Link Operation (MLO) technology, Wi-Fi 7 can use multiple routes simultaneously, ensuring fast and stable connections even when one route is experiencing interference—ideal for gaming and video streaming.

As connectivity standards advance, network equipment and infrastructure are also being updated. While these changes may not be as noticeable as server upgrades, the demand for improved network equipment and infrastructure will continue into the future.

Powering data connections

Semiconductor performance must advance in tandem with upgrades in connectivity standards. Semiconductors are crucial for preventing issues like poor connections in elevators or varying signal strength depending on location. Semiconductors amplify signals, ensuring wider coverage and distortion-free transmission.

While semiconductors remain essential, the market for semiconductor components supporting 5G is expected to grow relatively slowly compared to data centers. 5G infrastructure is largely mature in many countries, leading telecom companies to focus their investments on data centers. Therefore, growth in semiconductors for telecom equipment is likely to remain modest through 2030.

On the other hand, with increasing data traffic and the widespread adoption of AI in enterprises, demand for switches, routers, and intelligent network interface cards (NICs) that support data center cloud service operations is rising significantly. This shift is driving growth in the data center network equipment, local area network (LAN), and wide area network (WAN) markets. Consequently, semiconductors for these areas are expected to experience strong growth through 2030.

The growing volume of data requires more advanced and complex network equipment. Consequently, demand for application-specific integrated circuits (ASICs) and field-programmable gate arrays (FPGAs) is increasing, and more telecom equipment companies are likely to develop their own chips.

The Telecom Market's Shift to Gallium Nitride RF Chips

Radio frequency (RF) chips perform the critical function of amplifying wireless transmission signals. The advent of millimeter-wave 5G requires semiconductors capable of handling high-frequency ranges, and gallium nitride (GaN) and gallium arsenide (GaAs) meet this need. Furthermore, unlike GaAs, GaN can handle both high power and high frequencies, making it well-suited to the stringent specifications required for base stations. The advantages of GaN semiconductors are expected to be fully realized in sectors such as base stations, defense, and aerospace, where reliable communications are crucial.

Although growth may slow as 5G telecommunications equipment technology matures in developed countries, significant global growth potential remains through 2025, driven by continued demand for upgrades to existing base stations. Although GaN RF chips are more expensive than silicon-based options, the market for GaN RF chips is expected to continue growing. GaN already accounts for over half of RF chips in the telecommunications equipment market and is projected to capture up to 90% of the market share in the future. However, this does not mean that silicon-based RF chips will be completely replaced. Base stations typically amplify signals through multiple stages, making silicon-based chips a cost-effective option for low-frequency applications.

Semiconductor demand by application in 2030

Artificial Intelligence and Connectivity

The increasing complexity of AI models and the expansion of datasets are driving a significant increase in demand for high-speed data transmission within data centers and across networks. Consequently, demand for servers capable of managing this growing data traffic is increasing significantly, driving market growth.

Demand for accelerator cards, including AI accelerators, is surging due to increased demand for advanced servers, improved AI capabilities, and customized designs. Furthermore, demand for enterprise LANs will grow to handle faster data traffic. As communication standards evolve, the proportion of chips in the total cost of mobile base stations and fixed access equipment is expected to increase.

Concurrently, the market size for enterprise wide area networks, wireless local area networks (WLANs), and backbone infrastructure is steadily expanding due to rising data traffic and communication demands. However, the proportion of semiconductor costs is unlikely to increase significantly, as major investments in telecommunications infrastructure have already been made.

Home appliances

Despite a relatively saturated home appliance market, AI and IoT technologies are making appliances smarter and providing new consumer experiences. Furthermore, new home appliance categories such as augmented reality (AR)/virtual reality (VR) and wearables are gaining market acceptance.

The growth of AI-powered home appliances is likely to significantly increase demand for semiconductors such as AI processors and power management integrated circuits (PMICs), which achieve both energy efficiency and a personalized experience. Wearable devices for gaming and healthcare are also driving growth in the semiconductor markets for sensors, connectivity integrated circuits, and processing units. Finally, the expansion of the IoT is likely to spur demand for connectivity integrated circuits that support diverse communication protocols between electronic devices.

Semiconductors can serve as the foundation for the continued evolution of home appliances, fundamentally changing the "smart experience" in our homes.

Traditional home appliances and new products

While the home appliance market is relatively mature, the integration of IoT and AI technologies with traditional appliances like refrigerators can accelerate consumers' replacement cycles. Meanwhile, innovative products like augmented reality/virtual reality headsets and personal robots are beginning to penetrate the home appliance market.

Smart, powerful, and energy-efficient: the next wave of home appliances

The rise of intelligent experiences in smartphones, PCs, and cars is also raising consumer expectations for home experiences—not just in emerging appliances like augmented reality and virtual reality devices. While shipments of traditional home appliances haven't seen exponential growth, AI capabilities are being integrated into TVs, robot vacuums, and refrigerators, driving demand for AI processors in the appliance market.

In addition to performance, energy efficiency is crucial. AI can also improve the energy efficiency of home appliances. This is particularly important as energy efficiency standards for appliances become increasingly stringent, including regulations on standby power. This growing focus on energy efficiency is expected to drive demand for application processors designed for AI workloads, as well as power management integrated circuits (PMICs) and compact battery management ICs that optimize device power consumption.

The rise of the hyper-connected home experience

Home appliances are more connected than ever before, setting a new standard for smart experiences. Devices such as washing machines, refrigerators, lights, AI speakers, and robot vacuums now communicate with each other, creating a seamless smart home experience. The launch of the Matter smart home standard in 2022 further accelerated this trend, enabling devices from different manufacturers to communicate seamlessly.

As IoT connectivity becomes prevalent in home appliances, more devices will be equipped with connectivity integrated circuits. Furthermore, these chips—whether standalone system-on-chips or integrated into application processors (APs)—are evolving to support multiple communication protocols.

To achieve a smooth communication experience, devices utilize multiple channels depending on the situation. Matter supports Bluetooth, Wi-Fi, and a direct device-to-device communication protocol called Thread. For example, Wi-Fi is suitable for large, high-speed data transfers and is ideal for smart TVs and smart refrigerators with displays. Bluetooth is used for short-range communication, while Thread, with its energy efficiency, is suitable for battery-powered devices or smart speakers that need to connect directly to other appliances.

In summary, as home appliances increasingly adopt and expand their smart IoT capabilities, demand for connectivity integrated circuits is expected to rise.

Bridging the gap between reality and virtuality

How many devices do you wear now? From headphones to smartwatches, fitness bands, augmented reality/virtual reality devices, and healthcare equipment, wearables have become an integral part of our lives. Consequently, the demand for sensors is rapidly increasing. AR/VR devices may utilize a variety of sensors to track eyes and movements for user interaction, while cameras and microphones are used to provide augmented reality experiences. Wearable medical devices also use multiple sensors to monitor health, movement, and the environment. Inertial sensors track your movement speed, while magnetic field sensors support analysis by sensing your body's movements. New sensors, such as non-invasive blood glucose sensors that use saliva or photoacoustic methods, are likely to emerge.

The data collected from these sensors can often be noisy and irregular. Biosignals fluctuate with movement, and external interference from noise and electromagnetic fields can compromise accuracy.

To address these challenges, the semiconductor industry is also focusing on developing advanced processors and wearable device-specific system-on-chips to enable more efficient sensor data processing and improve overall device performance.

Semiconductor demand by application in 2030

Intelligence and the Internet of Things

Home appliances are becoming increasingly smart and connected. Televisions, the largest-selling category, now feature AI-driven image and sound enhancement, smart home controls, and personalized content recommendations, all of which are driving demand for advanced semiconductors.

Major appliances and digital set-top boxes are also increasing their semiconductor usage due to the addition of smart features and connectivity options, while wireless headphones and digital cameras are using more semiconductors due to their expanded functionality and sophistication.

On the other hand, smart speakers and consumer drones face less pressure to innovate, and since the majority of their costs lie in non-semiconductor components, their semiconductor demand growth is relatively modest.

Computing devices

Despite the maturing smartphone and PC markets, their value proposition is shifting towards high-performance models that redefine the user experience. The emergence of AI-driven applications, from advanced photography and gaming to real-time AI assistants, will reignite market growth and usher in a new generation of AI PCs and smartphones.

With the increasing demand for AI-integrated computing devices, the adoption of neural network processors (NPUs) is expected to accelerate, complementing advances in graphics processing units (GPUs), central processing units (CPUs), and image signal processors (ISPs) in application processors. Meanwhile, low-power double data rate (LPDDR) memory technology is likely to continue to advance, enabling improved performance, miniaturization, and greater energy efficiency for the next generation of PCs and smartphones.

With the proliferation of AI applications and the growing demand for higher-resolution displays, powerful computing power, and greater storage capacity, PCs and smartphones are likely to continue driving growth in the semiconductor industry.

Artificial intelligence drives further growth in computing devices

Compared to other applications like automotive and data centers, the smartphone and PC markets are relatively saturated.

However, as AI permeates our daily lives, there's a trend toward running AI services directly on personal devices. This shift is driving demand for AI-powered PCs and smartphones, as users seek devices that can seamlessly handle advanced AI applications.

The increasing integration of AI capabilities, from virtual assistants to on-device machine learning tasks, is revitalizing the market and creating new growth opportunities for device manufacturers and semiconductor companies.

On-device AI unlocks the next potential of smartphones and PCs

Smartphones and PCs have consistently driven demand for advanced chips using advanced process technologies. They must process large amounts of data with low latency while maintaining portability, convenience (through thin and light designs), and long battery life. Therefore, the performance of application processors (APs), which integrate high-performance computing units such as central processing units (CPUs), graphics processors (GPUs), and image signal processors (ISPs), has been a key competitive factor in the smartphone and PC markets.

A trend further driving device performance is the shift toward "edge AI." Previously, these tasks were handled by CPUs or GPUs, but the increasing complexity of AI models and the increase in sensitive data have led to the adoption of neural network processing units (NPUs) within devices. NPUs are dedicated AI processing cores integrated into application processors/system-on-chips (SoCs), enabling faster and more secure data processing. This trend is expected to drive semiconductor market growth, particularly in high-end product lines.

With faster and more secure NPUs and advanced on-device AI, smartphones and PCs can deliver seamless AI experiences. Smartphones may be able to summarize conversations in real time and instantly enhance photos without the need for external applications. PCs can enable AI-powered noise reduction, make video calls clearer, and provide real-time subtitle translation without lag.

Demand for AI-enabled smartphones and PCs is already growing, and neural network processors (NPUs) and edge AI are likely to continue driving the chip market for these industries.

The AI Behind PCs and Smartphones

With processor advancements, high-performance DRAM is essential to supporting faster data transfer and seamless processing. This demand is further driven by on-device AI, where real-time AI processing requires efficient storage solutions to manage high data loads while maintaining energy efficiency.

While high-bandwidth memory (HBM) is synonymous with high-performance storage, its power consumption limits its use in smartphones and PCs, where battery life is paramount.

Low-power double data rate (LPDDR) DRAM addresses this challenge by balancing high-performance processing with high energy efficiency, making it key to next-generation smartphones and AI-driven computing.

Each generation of LPDDR reduces power consumption by 30%-40% by lowering operating voltage, while design and process advancements within the same generation can deliver an additional 10%-30% savings. The upgrade cycle for both inter-generational transitions and intra-generational upgrades has lengthened from one to two years to approximately three years. LPDDR6, expected to launch in 2026, is expected to reduce power consumption by approximately 50% compared to LPDDR5, with further improvements expected by 2030. As AI workloads expand and energy efficiency remains a priority, LPDDR is likely to continue driving DRAM market growth, enabling high-performance, energy-efficient computing in mobile and PC devices.

Transforming Amateur into Professional: Image Signal Processor

Photography once required meticulous manual setup, but waves of automation—first with point-and-shoot film cameras, then with digital point-and-shoot cameras, and now with smartphones—have made high-quality imaging effortless.

At the heart of this process lies the camera sensor and image signal processor (ISP). The sensor acts like the "eye," receiving light and converting it into electrical signals. The ISP acts as the brain, analyzing and processing these signals in real time to improve the final image. High-performance smartphone cameras rely on the synergy between high-resolution sensors, multiple lenses, and, most importantly, a powerful image signal processor (ISP).

As camera features become more advanced, consumer expectations for better photos are rising. To meet this demand, smartphones now feature multi-camera module setups—such as triple or quad cameras—and higher-resolution sensors, which in turn require faster image signal processors (ISPs). Naturally, the demand for higher pixel counts in lenses or camera sensors increases, but utilizing these resources and achieving high-quality images is the role of the ISP.

At the same time, narrower bezels and higher screen-to-body ratios leave less internal space, making stacked lenses, folded periscope optics, and low-power ISP IP blocks crucial. As a result, advances in miniaturized camera modules and highly advanced on-chip ISPs are becoming a new growth driver for smartphone semiconductors, enabling future devices to capture professional-grade images in the palm of your hand.

Semiconductor demand by application in 2030

On-device AI

As smartphone functionality continues to expand, they are no longer simply communications devices; their role as computing devices is strengthening. Furthermore, smartphone semiconductor demand is currently experiencing stronger growth than other computing devices, such as PCs and laptops.

High-end smartphones, due to lower peripheral costs (such as camera lenses and displays) and more affordable prices, are experiencing strong demand. In contrast, demand for low-end models is relatively weak.

Recently, with the increasing incorporation of AI capabilities in computing devices, the proportion of semiconductor costs in cost of goods sold (COGS) has rapidly increased, particularly in laptops and smartphones. The more aggressive integration of AI capabilities in laptops compared to desktops has resulted in a relatively larger proportion of semiconductor costs in laptops.

On the other hand, demand for applications such as smart cards and external removable storage is relatively weak.

Semiconductor demand intensity by application in 2030

...

...

Source: Semiconductor Industry Observer

View more at EASELINK