In recent times, the global Consumer Electronics Show (CES) held in Las Vegas has become a melting pot of technology innovation, showcasing over 4,000 companies from 160 countries and regionsAmong them, more than 1,300 Chinese companies and 309 Fortune 500 firms like Nvidia, AMD, and Lenovo were present, making their mark in this critical tech arena.

This year, tech giants AMD, NVIDIA, and Intel heralded a new wave of "smart hardware" driven by artificial intelligence (AI). Notably, their innovations revolved around AI technology, signaling a clear trend: edge AI is not just a buzzword but a core component of modern technology that is swiftly integrating into everyday life.

In fact, in recent years, almost every hardware component in the market has started to incorporate AI functionalities

It is almost a given that without AI, there is no modern deviceConversely, the urge for advanced AI models depends on robust hardware; any new product must address three fundamental questions: what is the application context, how will it be accepted by the market, and how can it generate profit?

AI Integration in Personal Computers

Among key highlights was the emergence of the AIPC, or Artificial Intelligence Personal ComputerThis new category of PC integrates AI technology through the combination of NPU (Neural Processing Unit), CPU, and GPU, fundamentally altering user experience and significantly augmenting productivity and creativity

• OPEC+ Delay: What Lies Ahead for Oil Prices?
• Gold Weakens and Declines
• Gold & Silver: Volatile Ahead of Non-Farm Payrolls
• Hedging Options as Bitcoin Breaks $100K Barrier
• Coal Power's Low-Carbon Shift: External Aid & Internal Upgrades

The AIPC can be utilized across various scenarios, including graphic visualization, semantic understanding, and intelligent interaction, transforming traditional personal computers into AI assistants that act like a "second brain" for users.

The AIPC concept first surfaced in September 2023, but its real explosion is anticipated in 2024, marking it as the "Year of AIPC." At CES this year, top manufacturers launched an array of new products aligned with this trend.

On May 20, 2024, Microsoft presented Windows 11 AIPC, announcing the full integration of its AI assistant Copilot into the Windows system, utilizing OpenAI's GPT-4o model

Microsoft defined the AIPC standard as "Copilot + PC," emphasizing the need for integrated NPU, computing power exceeding 40 TOPS, and a minimum of 16GB of memory to guarantee that the computer can independently perform AI operations without relying on external servers.

Currently, leading chip manufacturers like Intel, AMD, Qualcomm, and NVIDIA continue to release high-performance processors designed for AIPC.

Intel unveiled its new Core Ultra (2nd generation) processors, which include the latest Intel Core Ultra 200HX and 200H series mobile processors, along with the Intel Core Ultra 200V series mobile processors based on the Intel vPro platform, and expanded offerings in the desktop Core Ultra 200S series designed for edge computing

According to Intel, the performance gains in the 200H series compared to its predecessor are approximately 17% for single-threaded tasks, 20% for multi-threaded tasks, with an integrated graphics performance increase of about 22%, and AI computing power reaching up to 99 TOPS.

AMD also announced its new Ryzen AIMax series processors, addressing the high-performance needs of premium ultrabooks, alongside the newly architected Ryzen AI300 series based on the "Zen5" architectureAMD is keen to leverage its "Zen 4" architecture with the release of Ryzen 200 series processors suitable for everyday use and expanding its business AIPC product line by integrating AMDPRO technology into the Ryzen AIMax, AI300, and AI200 series processors

The AIMax processors are bolstered by NPU capabilities with up to 50 TOPS of AI processing power, positioning them as formidable engines for the next generation of AI computers.

NVIDIA was equally competitive at CES, as CEO Jensen Huang unveiled the RTX 50 series graphics cards for PCs, reinforcing the power supply for AIPCMoreover, NVIDIA made a significant entry into the AIPC domain by launching a compact supercomputer, Project Digits, enabling more developers to utilize large AI models on desktop platformsThe new GeForce RTX 50 series boasts the NVIDIA Blackwell architecture, fifth-generation Tensor Cores, and fourth-generation RT Cores, with notable enhancements in AI rendering, including neural network shaders, digital human technology, geometry, and lighting.

Highlighting NVIDIA's innovations, the Project Digits supercomputer features the ground-breaking GB10 super chip, hailed as the smallest AI supercomputer to run 200 billion parameter models, a design collaboration with MediaTek

The GB10 is equipped with Blackwell GPU and connects through NVLink-C2C interconnect and a high-performance Grace CPU, utilizing ARM architecture with 20 efficient cores.

Qualcomm previously launched the Snapdragon X Elite and Snapdragon X Plus processors targeting PC products, yet devices using the Snapdragon X series accounted for less than 1.5% of the overall Windows PC marketIn a surprising move, Qualcomm canceled its developer kit last OctoberHowever, at CES, Qualcomm made a triumphant return with the launch of its new system-on-chip, Snapdragon X, aimed at the mid-range PC marketUtilizing a 4nm processing technology, this chip is designed to deliver long battery life and robust performance, accommodating the latest AI software and enabling PCs priced as low as $600.

With the backing of chip manufacturers, brands like Lenovo, HP, and Dell have ramped up research and production investments in the AIPC sector, releasing related computer products

At CES, AIPC terminal manufacturers also launched new devices.

Lenovo showcased the YOGA Air 14 Aura AI, ThinkPad X9, tablets featuring local AI, handheld devices, and the latest gaming laptops powered by NVIDIA RTX5090 graphics cards among a range of new productsDell introduced a new AIPC product range designed for both personal and professional computing, adopting a naming convention similar to Apple's with Dell, Dell Pro, and Dell Pro MaxASUS also exhibited a product touted as the "lightest Copilot + PC."

Furthermore, outside of the Windows ecosystem, Apple plans to launch an AI suite in June 2024, featuring its first generative AI large model, Apple Intelligence, which aims to capture a slice of the AIPC market.

AI Integration in Smartphones

In addition to AIPC developments, smartphone manufacturers are riding the wave of AI integration

The year 2024 is being labeled as the "Year of AI Smartphones."

According to Li Weimin, deputy chief analyst at the Strategic Development Research Institute of China Telecom, advancements in natural language processing have equipped voice assistants with "high IQ," allowing precise interpretation of user intentions, while multimodal interactions combining voice, gestures, and facial expressions make operations more intuitive and engagingHe remarked that AI will undoubtedly be the catalyst in the competitive smartphone market, motivating brands to boost AI-related R&D, as those who can thrive in this domain are likely to ascend to the proverbial throne.

During Apple's fall product launch event in 2024, CEO Tim Cook demonstrated the Apple Intelligence feature, with the new iPhone 16 series being heralded as Apple's first true AI smartphone

On October 22, Huawei officially revealed HarmonyOS NEXT, deeply integrating AI with the OS to deliver a novel Harmony Intelligence native experience, initiating an era where AI models redefine the operating system experienceWith its robust AI framework, the PanGu large model's Xiao Yi intelligence agent melds seamlessly with the system AI navigation bar, becoming an ever-present intelligent assistant on the interface.

Honor also launched a smartphone with a new AI system during its October product announcementTo illustrate the AI capabilities vividly, Honor's CEO Zhao Ming utilized the AI features to order 2,000 cups of coffee for guestsAdditionally, Xiaomi debuted its "Super Xiao Ai," Vivo introduced "Blue Heart Large Model," and OPPO unveiled "AndesGPT."

By the end of 2024, sources indicated Xiaomi is actively constructing its GPU cluster, enhancing investments into AI models

The Xiaomi large model team had already deployed 6,500 GPU resources at its inception, demonstrating a strong commitment to AI hardwareLast May, Xiaomi announced that its large language model, MiLM, had officially completed large model registrationXiaomi stated that the large model will gradually be implemented across its automotive, mobile, and smart home products, facilitating a multi-device collaboration strategy that empowers its overarching “vehicle-home ecosystem” strategy.

Beyond the terminal manufacturers, contributions from MediaTek and Qualcomm in the AI sector are equally commendable, with new products like the Dimensity 9400 and Snapdragon 8 Elite giving robust technological support for the rapid evolution of AI smartphones.

Integrating MediaTek’s Dimensity AI agentic engine and an eighth-generation AI processor, NPU 890, the Dimensity 9400 yields a significant performance boost for traditional AI applications and elevates the technology to a new height of autonomous perception, reasoning, and collaborative action

This means the Dimensity 9400 can fully enable edge AI, swiftly upgrading conventional applications across diverse sectors including text processing, image processing, and music creation.

The Snapdragon 8 Elite features an upgraded Hexagon NPU, providing innovative multimodal model processing capabilities, allowing AI computations to work harmoniously with computer vision, thus significantly enhancing the efficiency of AI tasksFor instance, through LMM (multimodal models), the NPU can simultaneously process voice commands and visual content, leading to faster response timesThis high processing efficiency positions the Snapdragon 8 Elite as favorable across various intelligent application scenarios.

Notably, the trend observed within development conferences for various smartphone brands has shifted from merely optimizing existing phone functions to focusing on AI agents at the core, leveraging localized edge models to offer more personalized services and new experiences tailored for mid-to-high-end consumers

The deep integration of AI capabilities with their operating systems has become a pivotal characteristic of flagship smartphones and a critical opportunity for companies to gain an edge in brand and market competition.

AI Fuels Intelligence in Various Hardware Domains

Smart hardware represents a powerful technological concept following smartphones, transforming traditional equipment into intelligent solutions through the integration of hardware and softwareThis transformation enables hardware to connect, download internet services, and establish a "cloud + edge" architecture, enhancing added value through big data.

Building on this, smart hardware has expanded from wearable devices to smart TVs, smart homes, smart vehicles, healthcare solutions, intelligent toys, and robotics

With the infusion of AI, various new products like MCUs, MPUs, and SOCs are increasingly incorporating AI functionalities, further propelling advancements in smart hardware.

In 2024, ByteDance will launch its first AI smart headset, Ola Friend, which integrates its Doubao large model and connects with the Doubao appThis AI smart hardware becomes ByteDance's exploration into AI scenarios, providing users with direct assistance in various life contexts.

At CES 2025, Samsung introduced the industry's first display that integrates AI directly into the device, while LG unveiled new smart refrigerators and microwaves wired for AI-driven smart control and energy management, automatically adjusting temperature and operation modes according to user habits.

Recently, Google and Synaptics announced a collaboration to explore the best solutions for contextual awareness computing multimodal processing in the realm of edge AI for the Internet of Things (IoT). This partnership aims to integrate Google’s MLIR-compliant ML kernel with Synaptics Astra hardware, accelerating the development of AI IoT devices that support processing for visual, imagery, audio, and sound across various applications including wearables, home appliances, entertainment systems, embedded hubs, monitoring systems, automotive, and enterprise solutions.

The "AI+" phenomenon is not merely a reflection of technological progress but illustrates how technology can play a crucial role in enhancing quality of life and driving societal progression

Through the synergy of smart hardware and AI technologies, future living is bound to become more convenient, efficient, and intelligent.

AI Injects New Vitality into Consumer Markets and Drives Business Growth

The integration of AI has reinvigorated the consumer markets for PCs, smartphones, and smart hardware.

In the PC market, IDC forecasts that the AIPC developments will end the negative growth trend in China, predicting stable growth in the coming five yearsThe total size of the PC market is projected to rise from 39 million units in 2023 to over 50 million units by 2027, reflecting an increase of nearly 28%. The proportion of new AIPC installations in China's PC market is expected to rapidly increase, reaching 85% by 2027, establishing AIPC as a mainstream component of the market.

In the smartphone sector, analysts predict that the implementation of AI applications will spark a new replacement cycle by 2025, with the global smartphone market expected to continue a gentle recovery

CEB International's report foresees that the arrival of AI applications on smartphones will further accelerate this replacement cycle, estimating a 3% year-on-year growth in global smartphone shipments to 1.25 billion units in 2025.

Fan Zhiyuan, chief analyst at Guojin Securities, believes that looking ahead to 2025, AI applications will keep driving the semiconductor cycle upwardsThe applications spurred by generative AI are expected to become mainstream within this AI wave, primarily presenting as "AI+X" solutions, encompassing AI smartphones, AIPCs, AI glasses, AI headphones, AI speakers, AI gaming, AI education, AI companionship, robotics, and autonomous drivingThe upgrade and innovation of demand for these endpoints will boost the need for chips, fostering substantial expansion within the overall semiconductor market.

Against the backdrop of this new development landscape, effectively applying AI has become a central topic of discussion