Gate News reports that in March 2026, the price of Bittensor (TAO) doubled, rising from around $180 to over $330, closely linked to the rapid expansion of its subnet ecosystem. Over the past year, the number of subnets in this decentralized artificial intelligence network has grown from about 80 to more than 120, pushing the total market value of the subnet sector to over $1.4 billion.
In this round of increases, several core subnets have become the focus of the market. Among them, Subnet 3 (Templar) trained the Covenant-72B model in a permissionless manner and was mentioned by Jensen Huang in a public discussion, raising market attention towards Bittensor’s technological roadmap. Meanwhile, the activity of Subnet Chutes (SN64) has significantly increased, with on-chain data indicating that the network’s revenue has set a record, achieving daily earnings of $22,000, showing its preliminary commercialization capability.
On the infrastructure front, Manifold Labs and Intel jointly released a technical white paper on trusted execution environments, proposing a secure execution of decentralized computing power through Intel TDX and cryptographic virtual machines. This solution emphasizes data privacy, cost transparency, and enterprise-level security capabilities, providing a new pathway for the openness of AI computing resources and further enhancing institutional interest in the Bittensor ecosystem.
From a market structure perspective, most subnet tokens have currently achieved positive returns, with funds shifting from single-token speculation to a layout within the subnet ecosystem. As the integration of AI model training, computing power markets, and blockchain incentive mechanisms gradually materializes, the logic behind TAO’s price increase is shifting from expectation-driven to fundamentals-supported.
In the context of the accelerated fusion of artificial intelligence and cryptocurrency infrastructure, the continued expansion of the Bittensor ecosystem may become an important variable in the next round of narratives.