Summary of Key Points
Brain-inspired intelligence is a cutting-edge technology that combines neuroscience and information science, capable of overcoming the limitations of traditional artificial intelligence’s “Von Neumann architecture” (the separation of computation and storage, which leads to high power consumption and slow data transfer). It represents a crucial path for the next generation of general-purpose AI. The global brain-inspired intelligence market is experiencing rapid growth, with revenues expected to rise from $28 million in 2024 to $8.352 billion by 2034, with the brain-inspired computing segment leading the growth at a compound annual rate of 67.3%. Countries like the United States, Europe, and Japan each have their unique approaches to this field; China leads in terms of the number of research papers and patents but faces challenges in the software aspect. The industry has already established a three-tier structure consisting of core technologies, integrated services, and industry applications. Future development will follow three phases: short-term validation, mid-term ecosystem building, and long-term breakthroughs in mechanisms. Partnerships are recommended to promote the industry’s advancement through standardization, technological innovation, and ecosystem development.
I. Brain-inspired Intelligence: The Solution to Traditional AI’s Limitations
Why is traditional AI becoming increasingly inefficient? It uses the Von Neumann architecture, which divides computation (like the kitchen) from storage (like the warehouse). This leads to unnecessary data transfers and energy inefficiencies, especially with large datasets. The issue is even more pronounced in the era of large models, where training a model can consume millions of dollars in electricity and result in slow real-time responses.
Brain-inspired intelligence aims to mimic the human brain by integrating computation and storage (similar to how neurons process information), using event-driven approaches (only processing relevant data) and sparse coding (using minimal resources for complex tasks). This approach enables energy efficiency, faster responses, and the ability to learn continuously. In simple terms, it makes AI both smarter and more energy-efficient, breaking through current limitations in computing power and energy efficiency.
II. Explosive Market Growth and Huge Potential in Sub-sectors
According to reports:
- The global brain-inspired intelligence market will grow from $28 million in 2024 to $822 million by 2029 (a nearly 30-fold increase) and reach $8.352 billion by 2034 (a 300-fold increase).
- Among the sub-sectors, brain-inspired computing (e.g., brain-inspired chips) is growing at a compound annual rate of 67.3%, while brain-inspired perception (e.g., sensors mimicking human eyes/ears) is growing at 47.9%.
The rapid growth is driven by the need for brain-inspired technologies in scenarios that traditional AI cannot address, such as low-power smart watches and real-time responsive robots. Both companies and investors are actively investing in this area.
III. Diverse Global Approaches, with China Leading in Research and Patents
Major economies have different strategies:
- United States: Focusing on a platform-based approach, developing not only chips but also supporting tools like programming frameworks and simulation software, with long-term support from organizations like DARPA (Defense Advanced Research Projects Agency) and NSF (National Science Foundation).
- Europe: Utilizing large-scale research initiatives, such as the Human Brain Project (HBP) and the EBRAINS virtual platform, to advance both brain science research and system validation.
- Japan: Emphasizing a combination of hardware and applications, with efforts in bio-inspired sensors and robot integration, supported by national brain-related programs.
China’s Advantages:
- Research Papers: 4,775 published from 2016 to 2026, accounting for 41.4% of the global total (the highest share), with this proportion expected to continue increasing.
- Patents: From 823 in 2016 to 19,212 in 2025 (a 22-fold increase), exceeding 50% of the global total.
- Policy Support: The government has launched major projects on brain science and brain-inspired research, with local governments in Beijing, Shanghai, Hefei, and other cities implementing policies to support technological development and industry clustering.
However, China’s weakness lies in the software layer, as it lacks a universal underlying platform like NVIDIA’s CUDA. This makes it difficult for developers to utilize hardware effectively, increasing the cost of industry applications.
IV. The Formation of a Three-tier Industry Structure and Emerging Downstream Opportunities
The brain-inspired intelligence industry consists of three main layers:
- Upstream Core Technologies: Brain-inspired chips, sensing devices, and other hardware (e.g., neurons-like chips), which are highly sought after by investors due to their technical complexity.
- Midstream Integrated Services: Combining hardware and software into complete solutions, a process that is still in progress.
- Downstream Applications: Smart robots, low-altitude economies (drones), connected vehicles, smart cities, and other applications.
The potential for downstream applications is significant, as brain-inspired technologies offer advantages such as low power consumption and high real-time performance. Investors recognize the importance of addressing real-world needs; for example, the director of the Shanghai Zuquan Innovation Transformation Research Institute suggests that investments should focus on practical use cases to drive upstream research and development.
V. A Three-step Pathforward for the Industry
The future development of the brain-inspired intelligence industry will follow these three phases:
1. Short-term (Validation Phase): Identifying irreplaceable use cases to demonstrate the technology’s value, such as low-power devices for wearable applications and IoT sensors.
2. Mid-term (Ecosystem Building Phase): Moving from individual technological breakthroughs to coordinated software and hardware development, establishing standards and frameworks to address software challenges.
3. Long-term (Mechanism Breakthrough Phase): Transforming the fundamental logic of intelligent computing to make brain-inspired intelligence a core component of general-purpose AI.
Recommendations from industry partnerships include:
- Establishing unified standards for measuring the performance of brain-inspired chips.
- Overcoming key technical barriers, especially in software platforms.
- Promoting open-source communities to make tools accessible to developers.
- Identifying benchmark applications for successful implementations, such as in robotics.
- Cultivating interdisciplinary talents with expertise in neuroscience, chip technology, and industry applications.
Brain-inspired intelligence is not a theoretical concept; it represents a practical solution to current AI challenges. In the coming years, we will see its application in various real-world scenarios, potentially leading to improvements in products like smart watches with longer battery life and faster responses.