Walrus, IO bring GPUs and decentralized storage to AI builders

AI app and agent devs can train and run their own custom models using cloud compute

article-image

Alex Hariyandi/Shutterstock and Adobe modified by Blockworks

share

This is a segment from The Drop newsletter. To read full editions, subscribe.


Decentralized cloud computing provider io.net, which has its own IO token, is working with the Walrus team to let startups train, run and store their own custom AI models.

IO offers a network of GPUs for AI training and fine-tuning, while Walrus is enabling AI model storage in the deal. The integration will be available as a pay-per-use offering, meaning builders are only charged for the amount of storage and computing power they use. 

This Bring Your Own Model (BYOM) offering lets AI agent devs or AI app builders develop and operate AI models without needing to set up their own data centers or hardware to do it.

IO says it has over 10,000 GPUs and CPUs globally.

Walrus and IOo’s BYOM offering will have to compete with other AI developer cloud services, though, like those from Bittensor, Lambda, Spheron, Akash, Gensyn, Vast AI, and Google’s Vertex products. 

“Traditional centralized cloud models are not only expensive — they come with significant privacy risks and limited composability options that are challenging for developers who prioritize decentralization,” said Rebecca Simmonds, Managing Executive at Walrus Foundation. 

“By leveraging our decentralized data storage solution, io.net will be able to provide the necessary compute power for advanced AI and ML development without any of the drawbacks of traditional models, making this a clear win for developers, users, and the entire Web3 industry,” the exec continued.

The internet can feel very centralized when giant cloud services experience outages, like what happened to Google last week (and impacted Cloudflare as well as a range of other apps and sites). That’s one obvious reason why centralized AI compute might not be ideal.

Walrus’s mainnet launched in March, with its main pitch being programmable, decentralized storage. The Walrus Foundation announced a raise of $140 million that month.

More broadly, the need for computing power for AI is expected to keep increasing annually. McKinsey researchers predict that data centers are going to need $6.7 trillion globally to keep up with demand by 2030 (though there’s a range of possible scenarios they modeled below).

AI compute is expected to need $5.2 trillion in investments to keep up with current demand in the next five years, while non-AI IT apps are expected to need much less than that at $1.5 trillion.

Walrus’s AI storage offering with IOo could help the Sui-based storage protocol see more adoption from AI builders and app devs, even from those outside of the crypto space. 

The Walrus mainnet currently has 4,167 TB in total storage capacity, with about 26% of that being used right now. The protocol has 103 total operators across 121 storage nodes.


Get the news in your inbox. Explore Blockworks newsletters:

Tags

Decoding crypto and the markets. Daily, with Byron Gilliam.

Upcoming Events

Old Billingsgate

Mon - Wed, October 13 - 15, 2025

Blockworks’ Digital Asset Summit (DAS) will feature conversations between the builders, allocators, and legislators who will shape the trajectory of the digital asset ecosystem in the US and abroad.

recent research

Research Report Templates (11).png

Research

We believe that Exponent is best positioned to dominate the Solana yield market – the fastest-growing sector in the ecosystem – with topline liquidity increasing at an annualized rate of nearly 600%. The founding team (ex-Squads, Kamino, Solana Foundation) cares deeply about product, security, design, and user experience and understands the foundation for building a great protocol. The team’s deep ties in Solana DeFi also present a significant strategic advantage, as evidenced by Exponent PTs being onboarded as collateral on Kamino, Drift, and Loopscale before its main competitor, RateX. In this regard, we view PT integrations in money markets as the most compelling avenue for Exponent to expand its market share and own its vertical.

article-image

Franklin Templeton’s Innovation Head Sandy Kaul says institutions will issue stablecoins and tokenized cash offerings to stay competitive

article-image

Surviving financial doomsday takes some preparation

article-image

Pump.fun’s launch partners include Kraken, Kucoin, Bitget, Bybit, Gate and MEXC

article-image

Monad Foundation’s Keone Hon said the team’s been “thinking about how to grow the onchain economy for some time”

article-image

Cuts to government R&D budgets may prove penny wise, bitcoin foolish

article-image

Still relatively little is known about the memecoin platform’s forthcoming token