Ocean Network, a decentralized compute platform for AI workloads, has announced the Beta launch of its peer-to-peer GPU orchestration layer. The system is designed to simplify access to high-performance hardware and reduce reliance on traditional cloud providers.
During the Beta phase, the company is sourcing GPUs through its 2025 partnership with Aethir, a distributed cloud computing provider. This gives users access to hardware such as NVIDIA H200 at competitive rates.
Ocean Network is offering $100 in compute credits to early users. The platform is доступна globally, with capacity scaling based on demand. In the future, it will allow individuals and companies to connect idle GPUs and earn revenue by providing compute resources.
A key component is the Ocean Orchestrator, an extension integrated into development environments like VS Code and Cursor. It allows users to select hardware, deploy jobs, and monitor execution in real time from a single interface.
The pricing model is based on actual usage. Payments are handled through an escrow system on Base, an Ethereum Layer 2 network for low-cost transactions. Funds are released only after tasks are completed.
For sensitive data, the platform uses a compute-to-data model. This means algorithms run where the data is stored, and only results are returned, without transferring raw data.
The Beta phase focuses on users running workloads, with plans to expand toward a broader decentralized compute marketplace.
Image: Freepik