第一财经

Qualcomm Announces 6G Research Progress: The Next Generation of Communication, Laying the Groundwork for AI

原文:高通公布6G研发进展:下一代通信,开始为AI铺路

Summary of the Key Points

The core message of this news is that the true breakthrough of 6G is not about faster internet speeds, but about the deep integration of AI into the communication system—ranging from end devices to the network itself, all the way to the business models of operators. Qualcomm predicts that 6G will be commercially available in 2029 and pre-commercially available in 2028, with three major changes:

1. The emergence of "agent terminals"—new devices that can perform tasks autonomously.

2. The transformation of 6G networks from mere data transmitters into "intelligent computing entities."

3. Operators shifting from selling data bandwidth to selling computing power. Qualcomm itself is also expanding beyond its role as a smartphone chip manufacturer to enter the field of computing power for various use cases.

1. 6G Terminals: Not Just Smartphones, but Intelligent Devices That Can Work on Their Own

In the past, when we thought of communication terminals, we meant smartphones. However, the 6G era could be completely different. Qualcomm suggests that "agent terminals" will emerge—these devices may not have screens and could be smart bracelets, rings, or even small accessories worn on the body. The biggest difference between these and current smart devices is that they will no longer passively follow your commands; instead, they will actively assist you in completing tasks. For example, your smart ring could automatically monitor your health data and contact a doctor if anything unusual is detected, or your smart bracelet could book a coffee or call a car service based on your schedule.

What's even more interesting is that companies that previously focused on cloud-based AI (such as those developing large models) might start manufacturing hardware. Since these agent terminals require AI capabilities, these cloud-based companies have a technological advantage and could potentially become new manufacturers of such devices.

2. 6G Networks: From "Data Transmitters" to "Intelligent Entities"

5G primarily aimed to make internet connections faster, but 6G integrates AI directly into the network. Qualcomm refers to this as the "computing continuum"—the combination of the computing power of terminals (like smart rings), edge servers (small data centers near users), and the cloud (large data centers). Tasks are automatically routed to the most suitable location for processing. For urgent matters (such as calling the police after a fall), the terminal can handle them on its own; for tasks that require more processing (such as video editing), edge computing power is used. Moreover, 6G networks can "perceive" their surroundings (such as identifying vehicles, pedestrians, traffic lights, and even drones) and use AI to analyze this information to automatically adjust network parameters (e.g., allocating more bandwidth to congested areas). Such networks, which can both process and understand their environment, are called "self-aware networks"—they are no longer just channels for data transfer but intelligent systems that can manage themselves.

3. Operators: Shifting from Selling Data Bandwidth to Selling Computing Power

In the 5G era, operators faced the challenge of investing heavily in infrastructure (base stations) while only earning profits from selling data bandwidth, resulting in low returns. With 6G, this situation is changing. Operators can now sell computing power services—instead of just data. For example, companies that need to run AI models can rent computing power from operators without having to purchase their own servers. They might also offer "token services" (digital entitlements that can be used to access specific AI services). Chinese operators have already started exploring this, with China Telecom entering the computing services market. For instance, a small company that needs AI for image recognition can purchase a computing power package from an operator, avoiding the expense and inconvenience of buying expensive GPUs and leveraging the 6G network.

4. Qualcomm's Ambition: From a Leader in Smartphone Chips to a Player in Computing Power for All Scenarios

Qualcomm has traditionally made its money from smartphone chips, but AI and 6G present new opportunities. The demand for AI computing has expanded beyond smartphones to include PCs, cars, robots, edge devices, and data centers. In June of this year, Qualcomm announced a roadmap for its data center business, aiming to generate over $15 billion in revenue from this segment by 2029. The reason for this success is that the "computing continuum" of 6G requires the coordination of computing power across all components, and Qualcomm's expertise in communications and chips gives it a competitive advantage in these emerging markets. For example, future biomimetic robots could use 6G to access cloud computing power without needing heavy hardware, and Qualcomm can provide the technology to connect these devices to the network.

Conclusion

AI is redefining everything about 6G. It is not just a speed upgrade of 5G but an "intelligent revolution" that combines AI with communication technology. It changes the form of communication devices, enables networks to think independently, and forces companies like Qualcomm to adapt their business models. For individuals, this means more intelligent devices that understand user needs, making internet access and AI services more convenient. For industries, the boundaries between communications and AI will blur, leading to new opportunities and competition.