Bittensor’s native token TAO has been one of crypto’s hottest movers this month, jumping roughly 73% over the past 30 days even as larger tokens posted more modest gains. The surge, analysts say, is tied to growing attention on decentralized AI after a high‑profile nod from NVIDIA’s CEO.
Why the excitement
NVIDIA CEO Jensen Huang recently acknowledged decentralized AI training as a practical approach after hearing about Bittensor’s latest technical milestone — a recognition that market analyst Alex Carchidi says helped validate the project’s core thesis and drew fresh investor interest. The milestone in question: Bittensor’s Templar subnet reportedly trained Covenant‑72B, a 72‑billion‑parameter large language model, via a decentralized collaboration of more than 70 contributors using commercially available hardware.
Why that matters
Training LLMs at this scale typically requires vast capital and tightly controlled data‑center infrastructure. Demonstrating that a 72B‑parameter model can be trained in a distributed environment gives real-world credibility to Bittensor’s idea that subnets can aggregate meaningful compute to produce economically valuable services.
Tokenomics and the catch
Carchidi also points to TAO’s supply design — which he says partially resembles Bitcoin’s issuance dynamics — as a potential long‑term upside for token holders if the network continues to deliver in‑demand services. But he warns of a major caveat: the subnets have not yet shown sustained external demand that converts into significant outside revenue.
Current economics underscore that concern. According to the analysis:
- The top subnet receives roughly $52 million in annualized TAO subsidies from the protocol.
- That subnet brings in at most about $2.4 million in outside revenue.
- Across the whole Bittensor network, demand‑side revenue is estimated between roughly $3 million and $15 million per year.
- By comparison, TAO’s market capitalization sits near $3.3 billion.
Carchidi calls this a “valuation mismatch,” warning TAO’s price could be vulnerable if subnets fail to scale revenue materially.
Technical picture
On the price charts, TAO was trading around $308 at the time of reporting. Immediate resistance sits near $315 — a level the token failed to consolidate above during last week’s rally. Crypto market analyst Ali Martinez previously flagged $315 as a critical pivot: if that level holds as support, the rally could extend toward $580. TAO’s inability to breach a mid‑term resistance at $378 contributed to the pullback to $308. The token remains nearly 60% below its all‑time high of $757.
Bottom line
Bittensor’s recent technical feat and a high‑profile acknowledgment from NVIDIA have boosted confidence that decentralized LLM training can work at scale. However, the network’s current revenue generation lags its token‑issuance subsidies, creating a potential valuation risk unless subnets can convert compute and model capabilities into meaningful external demand and revenue.
Read more AI-generated news on: undefined/news
