Walrus announced a deep integration with io.net to provide decentralized storage and computing capabilities for AI and machine learning applications.

According to 深潮 TechFlow, on June 19, the blockchain storage protocol Walrus built on Sui is deeply integrating with the decentralized computing network io.net. The integration of Walrus and io.net creates the "(Bring Your Own Model, BYOM) platform, where users can integrate custom AI models without relying on pre-fabricated or curated models. io.net provides the GPU clusters necessary for training and inference, while Walrus’s decentralized storage ensures proprietary models are secure and tamper-proof, building a trustworthy AI data storage and computing environment.

In addition to decentralized storage and computing capabilities, Walrus also supports privacy-preserving computation execution, allowing models stored in Walrus to be directly loaded onto the io.net GPU cluster for training or fine-tuning. The integrated solution provides top-notch encryption and access control, and employs a pay-as-you-go billing model to ensure that developers only pay for the storage and computing power they actually use.

This integration highlights that Walrus is becoming a key component of the decentralized AI technology stack. The BYOM platform of io.net (including model upload, computing, and billing) is currently undergoing internal testing and is expected to officially launch in a few weeks.