15th Five-Year Plan: Developing High-Performance AI Chips and Highly Available Foundational Software Stacks
nashnova research
China's Central Cyberspace Affairs Commission released the National Informatization 15th Five-Year Plan, elevating homegrown high-performance AI chips and a foundational software stack to national-strategy status — alongside quantum tech, humanoid robots, and brain-computer interfaces — effectively setting the scorecard for Chinese tech self-reliance through 2030.
Why are AI chips and the software stack singled out?
The plan calls for developing high-performance AI chips and a highly available foundational software stack, writing both hardware and software into the same national-level directive.
This means → policymakers see chips alone as insufficient; the base software that runs on them — OS, compilers, frameworks, essentially the chip's "interpreter layer" — must also be domestically controlled.
In plain terms = the chip is the engine, the software stack is the gearbox; missing either one and the car doesn't move. The plan demands both be built together.
What does the plan sketch for devices and robots?
On the device side, it targets highly intelligent smartphones, computers, and wearables — pushing AI capability down to end-user hardware.
In embodied intelligence — giving AI a physical body to perform real-world tasks — it explicitly names unmanned vehicle terminals and humanoid robots.
This means → the policy signal has shifted from "lab concept" to "industrialization." Humanoid robots now appear in five-year-plan-level language for the first time.
What mission does quantum technology receive?
The plan calls for accelerating R&D across three tracks: quantum computing, quantum communication, and quantum precision measurement.
It separately mandates work on quantum-safe security and post-quantum cryptography — encryption methods designed to withstand future quantum computers.
In plain terms = build the quantum spear and the quantum shield at the same time — offense and defense in parallel.
What does putting brain-computer interfaces in a national plan signal?
The plan backs R&D and application of brain-computer interface systems, covering signal acquisition, neural decoding — translating the brain's electrical signals into machine commands — disease treatment, and virtual-physical control.
This reflects a repositioning: policymakers now treat BCI not as a pure-research frontier but as a systems-level engineering push.
What is the ultimate test for this roadmap?
The plan itself states that whether AI chips and the foundational software stack achieve breakthroughs within the planning cycle is the core metric for judging the deployment's success.
This means → policymakers have set the hardest problem — chips and the software stack — as the plan's "exam score." No matter how cutting-edge the other tracks are, they all wait for this foundation to hold.
In plain terms = five years from now, the first question when grading this plan will be: can homegrown AI chips and their companion software actually be used in production?
市场有风险，内容仅供研究参考，不构成投资建议。