TMTPOST--Imagine having your own AI assistant that understands your needs, comprehends your emotions, and keeps track of all your work and life activities. What once seemed like science fiction is now gradually becoming reality.

In the past two years, generative AI has primarily developed in data centers and the cloud. However, with the increasing demand for energy efficiency, data privacy, and personalization from end users, AI technology for consumers has seen further advancements.

Gartner, an American technological research and consulting firm, predicts that by 2025, over 50% of enterprise-managed data will be created and processed outside data centers or the cloud. This marks the beginning of the era of generative AI for personal use.

Starting this year, personal computer (PC) manufacturers have been actively positioning themselves in the AI PC market. Canalys predicts that by 2027, the penetration rate of AI-powered PCs will reach 60%.

Following the first PC revolution of the graphical interface and the second revolution of the internet, the PC industry is now embarking on its third revolution—the era of generative AI.

According to Intel, over eight million AI PC processors have been shipped to date, and the number is expected to reach 40 million by the end of the year. By 2028, it is projected that 80% of PCs worldwide will be AI PCs, offering users a wide array of products and applications to choose from.

In this transition, energy efficiency, security, cost, and personalization have become key issues for AI models on the terminal side, particularly for PCs.

The CPU, as the brain of a PC, must keep pace with AI advancements to support the substantial computational demands. The industry is collectively addressing the challenge of transforming traditional CPUs into AI-supporting CPUs. According to Sun Wenjian, CEO of Cix Technology, constructing an efficient and ecological architecture is crucial in this transformative era for the CPU industry.

Unlike traditional CPUs, AI CPUs provide competitive solutions for various intelligent devices, including PCs, enhancing human-digital interactions, graphics rendering, multimedia data processing, and high-efficiency computational output. "End-side intelligent devices are diverse, including IoT, smartphones, computers, vehicles, and the metaverse, as well as future robots," said Sun. "These are all areas where AI CPUs can excel."

Creating new AI CPUs inevitably challenges existing ecosystems. Sun emphasizes that breaking these barriers requires highly experienced talent for complex chip design and substantial financial support.

For AI PCs, a functional CPU is just the beginning. Yan Yiqiang, chief researcher at Lenovo Group, noted that establishing a complete upstream and downstream ecosystem is essential. "Decoupling hardware, architecture, and models from applications is key," he emphasized.

While breaking the original ecosystem barriers is challenging, China's extensive AI PC market and complete industry chain offer significant advantages. Collaborative development with industry chain partners can further narrow the gap with world-leading players.

As AI continues to integrate into PCs, the industry is poised for profound changes, ushering in an era where personalized AI assistants become a common reality, fundamentally transforming how people interact with technology.