Participating operators
A path for people and organizations with suitable infrastructure to contribute, subject to defined requirements.
TAGII Gate / Research direction
Computing power. More places to contribute.
Exploring distributed infrastructure for authorized AI workloads, alongside the managed services that businesses need.
Research direction
More places to contribute computing power.
TAGII Gate explores distributed AI compute: participating operators could contribute suitable infrastructure to serve authorized workloads. Participation, workload placement and compensation require a tested operating model.
Authorized workloads
A workload needs an appropriate operator, data boundary and verification method. Private memory is not distributed to arbitrary machines.
Measured contribution
Capacity, reliability, settlement and operator responsibilities must be established before a paid network can open.
Compute and data need different decisions.
AI compute and data storage can work together, but they have different requirements for location, access, availability and recovery. The right boundary depends on the workload. Distributed compute does not mean every operator can see every person’s information.
This is a research and product direction. There is no open operator programme, published earnings schedule or general workload-placement offer on this page.