Bittensor (TAO) has emerged as one of the market\u2019s strongest performers this month, rallying roughly 73% over the past 30 days even as larger cryptocurrencies staged a more modest recovery.\u00a0<\/p>\n
Market analyst Alex Carchidi argues that a key catalyst was public recognition from a major tech figure. NVIDIA CEO Jensen Huang recently acknowledged decentralized AI training \u2014 the core use case Bittensor champions \u2014 as a practical approach after hearing about the project\u2019s latest technical milestone.\u00a0<\/p>\n
The analyst asserted <\/a>that the endorsement amplified investor interest in the Bittensor blockchain by validating the concept that training large language models (LLMs) can be accomplished outside centralized data centers.<\/p>\n Bittensor\u2019s most recent achievement centers on its Templar subnet, which reportedly trained Covenant\u201172B, a 72\u2011billion\u2011parameter LLM, through a decentralized collaboration involving more than 70 contributors using commercially available hardware.\u00a0<\/p>\n That accomplishment is noteworthy because large\u2011scale model training<\/a> normally requires substantial capital and tightly controlled infrastructure; proving such a model can be trained in a distributed manner lends credibility to Bittensor\u2019s thesis that subnets can marshal meaningful compute to build economically valuable services.<\/p>\n Carchidi also points to TAO\u2019s tokenomics <\/a>as a potential upside factor. The chain\u2019s supply dynamics partially resemble Bitcoin\u2019s (BTC), a design Cardichi finds appealing for long\u2011term appreciation if the network continues to produce in\u2011demand services.\u00a0<\/p>\n Still, he cautions that the project faces a critical shortcoming: so far, its subnets have not demonstrated robust, sustainable demand that translates into meaningful external revenue.<\/p>\nBittensor At Risk If Subnets Don\u2019t Scale<\/h2>\n