Node specialization

Nodes can specialize in select types of compute including AI Inference, ZK Proving, or TEE Code Execution, rather than servicing all operations, earning outsized rewards for performance.

Guardians

Granular firewall system for nodes to opt-in to servicing select types of execution, balancing security and performance preferences, while still participating in consensus.

Minimized re-execution

Minimized compute re-execution via Symphony. Nodes verify succinct proofs and participate in committees rather than re-execution, reducing bandwidth usage and improving network efficiency.

Native Infernet Integration

Use the expansive Infernet network as a compute job backend, delegating complex operations to specialized Infernet nodes. Scale compute offerings without requiring local resources yourself.