According to the latest forecast report from International Data Corporation (IDC), the shipment volume of artificial intelligence (AI) personal computers (PCs) is expected to experience significant growth in the coming years. Starting from nearly 50 million units in 2024, this number is projected to surpass 167 million units by 2027. These AI PCs refer specifically to personal computers equipped with specific system-on-a-chip (SoC) that have the capability to perform generative AI tasks locally. IDC further predicts that AI PCs will account for nearly 60% of the global PC shipment volume by the end of the forecast period, marking the transition of this technology from a niche field to the mainstream market. Tom Mainelli, Vice President of Devices and Consumer Research at IDC, stated, "As we enter a new year, the buzz around generative AI has reached unprecedented levels. The PC industry is quickly leveraging the potential benefits of moving AI capabilities from the cloud to the client. These advantages, including improved performance for faster user productivity, reduced inference costs, and assurance of privacy and security on the device, have sparked strong interest among IT decision-makers in AI PCs." In the past, running AI tasks locally on PCs primarily relied on central processing units (CPUs), graphics processing units (GPUs), or a combination of both. However, these chips were not optimized for AI and could potentially have negative impacts on PC performance and battery life. PC silicon manufacturers have now begun integrating neural processing units (NPUs) specifically designed for AI into their SoCs to execute these tasks more efficiently. According to IDC's classification, there are currently three types of NPU-supported AI PCs: 1. Hardware-supported AI PCs, which are equipped with an NPU that provides performance below 40 trillion operations per second (TOPS) and can typically enable specific AI functionalities to run locally in applications. Companies such as Qualcomm, Apple, AMD, and Intel have already introduced relevant products in this field. 2. Next-generation AI PCs, which include an NPU providing 40-60 TOPS performance and an AI-focused operating system (OS) that enables persistent and pervasive AI capabilities in both the OS and applications. These companies have also announced future chip plans in this category and expect to begin deliveries in 2024. Microsoft is expected to introduce significant updates and updated system specifications to Windows 11 to fully leverage these high-performance NPUs. 3. Advanced AI PCs, which are projected to offer NPU performance exceeding 60 TOPS. Although no silicon manufacturer has announced products in this category yet, IDC expects them to emerge in the next few years. There are three major technological advantages in shifting AI workloads from the cloud to the client: firstly, improving performance by eliminating the need for current AI workloads to travel back and forth to the cloud; secondly, enhancing privacy and security by storing data on the device instead of during transmission; and finally, reducing costs by minimizing access to expensive cloud resources. The report titled "Worldwide AI PC Forecast, 2023-2027" presents IDC's first forecast for the global AI PC market from 2023 to 2027, covering different NPU performance categories, processor architectures, product categories, and market segments.