SyncAI.news, a Varaisys broadcasting
Nvidia Is Fueling The Rise Of Alternative AI Networks
MM

Murtuza Merchant, Contributor

· 2 min read

World NewsForbes: Innovation

Nvidia Is Fueling The Rise Of Alternative AI Networks

As Nvidia prepares to report earnings Tuesday for the first time since Jensen Huang stepped down, decentralized AI networks are winning paying customers who can’t get GPUs from the big clouds even as the tokens behind them plunge. The divide between real demand and crypto speculation leaves investors with a simple question: Would these networks still have customers if their token rewards disappeared tomorrow?

Nvidia’s Rough Month Sets The Stage

Nvidia will report results for its fiscal third quarter on Nov. 17, with the company guiding to about $108 billion in revenue.

It is the chipmaker’s first report since Huang stepped down as CEO on Nov. 12, with finance chief Colette Kress serving as interim CEO. The stock was already more than 25% below its October record before his exit.

Crypto’s AI tokens have had an even harder time. They fell with the broader AI selloff on Nov. 5, and Bittensor’s TAO, the largest of them, trades roughly 70% below its peak.

Yet the businesses behind some of those tokens are growing. Bittensor’s subnets, the specialized markets that run on its network, are now generating about $70 million a year in revenue, according to an estimate by tracker SubConnect, up from about $35 million two months earlier. Customer spending on Lium, a GPU rental subnet, rose 87% in a month to $2.7 million.

MORE FOR YOU

Why Decentralized AI Revenue Is Rising While Tokens Fall

The split makes sense once you look at what customers actually buy. “Customers don’t buy tokens, they buy GPU hours, and those are priced in dollars. That’s why demand and token prices can move in opposite directions,” says Greg Osuri, founder of decentralized compute marketplace Akash Network.

The pull is access. The biggest cloud providers sell much of their new capacity to their largest customers through long contracts, leaving smaller buyers to wait.

The buyers are AI startups, app and agent developers and researchers running open-weight models, rather than frontier labs.

Even a niche can matter.

Original source

This story was published by Forbes: Innovation and written by Murtuza Merchant, Contributor. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on forbes.com

Similar News