Exploring the Cutting Edge: AI Applications in Action

The 2024 AI server market—encompassing AI Training and AI Inference—is projected to exceed 1.6 million units, growing at 40%, according to a report by TrendForce.

One notable shift in the Edge AI applications is the move towards AI PCs. This shift aims to decentralize AI server workloads and broaden the AI usage spectrum.

TrendForce expects AI PCs to meet Microsoft's benchmark of 40 TOPS (trillion operations per second) in computational power.

With new products meeting this threshold expected to ship in late 2024, significant growth is anticipated in 2025, especially following Intel's release of its Lunar Lake CPU by the end of 2024.

The AI PC market is currently driven by two key factors. Firstly, there is a high demand for terminal applications, mainly dominated by Microsoft through its Windows OS and Office suite. Microsoft is also planning to integrate Copilot into the next generation of Windows, making it a fundamental requirement for AI PCs.

Secondly, Intel, as a leading CPU manufacturer, is advocating for AI PCs that combine CPU, GPU, and NPU architectures to enable a variety of terminal AI applications.

Qualcomm's Snapdragon X Elite platform, set to be the first to meet Copilot standards, is expected to ship in the second half of 2024, delivering around 45 TOPS. AMD's Ryzen 8000 series (Strix Point) is also expected to meet these requirements. Intel's Meteor Lake, launched in December 2023 with a combined CPU+GPU+NPU power of 34 TOPS, falls short of Microsoft's standards. However, Intel's upcoming Lunar Lake might surpass the 40 TOPS threshold by the end of the year.

The competition between Qualcomm, Intel, and AMD in the AI PC market is set to intensify the competition between the x86 and Arm CPU architectures in the Edge AI market. Qualcomm's early compliance with Microsoft's requirements positions it to capture the initial wave of AI PC opportunities, presenting a challenge to the x86 camp as major PC OEMs like Dell, HPE, Lenovo, ASUS, and Acer develop Qualcomm CPU-equipped models in 2024.

Lastly, the CPUs meeting Microsoft's 40 TOPS requirement for NPUs include Qualcomm's Snapdragon X Elite, AMD's Strix Point, and Intel's Lunar Lake. These CPUs adopt LPDDR5x over the current mainstream DDR SO-DIMM modules, enabling faster data transmission. LPDDR5x reaches speeds of 7500–8533 Mbps, aiding AI PCs that require quicker language processing and responsiveness.

TrendForce forecasts that LPDDR will represent approximately 30–35% of PC DRAM demand this year, with future increases driven by the specifications set forth by AI PC CPU manufacturers, leading to a further rise in LPDDR adoption.