For providers

Prepare to provide infrastructure

A guide to the proposed provider role for datacenter operators, cloud providers, and HPC facilities. The network is not live; use this material to assess future participation.

Provider documentation Onboarding path
Server racks in a provider data centre.
Network and power connections inside a computing rack.

Why provide

What the protocol does for an operator

Monetize idle capacity

Turn unused racks, cloud headroom, or HPC cycles into revenue. The marketplace matches your capacity with tenant demand — publish list prices for direct orders, and set your own bid strategy for open ones.

Settlement without invoicing

Usage is metered, reported, and settled by the protocol. Payment releases from escrow automatically after the dispute window — no accounts receivable, no chasing customers.

Choose your counterparty requirements

VEID can provide a privacy-preserving trust signal. Providers may choose a clearly disclosed verification requirement for an individual offer; buyers can compare requirements and provider evidence before committing.

Reputation that compounds

Auditor-signed attributes, on-chain hardware benchmarks, and tenant reviews build a portable, tamper-evident track record that wins better-priced leases over time.

Economics

Usage in, payment out

Metered usage settles from escrow. Full money path lives in the canonical guides — not repeated here.

Usage is metered hourly as signed records, passes a 24-hour dispute window, then settles from lease escrow at the full agreed amount. Escrow & settlement → Provider economics →

Provider daemon

One daemon between the chain and your infrastructure

The provider daemon is the protocol's on-premises agent. You configure it; it handles the marketplace.

HPC

Supercomputing capacity is a first-class citizen

If you run SLURM, MOAB, or Open OnDemand, the daemon's HPC integration executes on-chain jobs on your cluster — with usage reporting, state recovery, audit logging, and retry built in.

slurm · moab · ood

Scheduler adapters

Native adapters speak to your existing controller — munge or JWT auth for SLURM, with per-partition configuration. No re-platforming of your cluster.

x/hpc

Job lifecycle service

On-chain HPC jobs are polled, dispatched, and tracked with configurable concurrency limits, timeouts, and crash-safe state recovery.

usage · security

Audit trail

Job events, security events, and usage reports are logged to a dedicated audit log, and usage batches flow into the same settlement pipeline as cloud workloads.

Requirements

What you need to bring

Onboarding

Readiness checklist

Onboarding

The planned provider workflow

  1. Verify your identity

    VEID is optional for general marketplace participation. Where a provider chooses a specific verification requirement for an offer, disclose it before the order so each side can make an informed choice.

  2. Register on-chain

    Create your provider record with attributes — region, hardware classes, certifications — via the x/provider module.

  3. Deploy the provider daemon

    Install and configure the daemon against your Kubernetes cluster or HPC scheduler; connect your chain key, and set list prices and bid strategy.

  4. Get audited & benchmarked

    Optional but valuable: auditor-signed attributes and published hardware benchmarks make your offers stand out to tenants that filter on them.

  5. Bid, serve, settle

    Direct orders match your list prices; the daemon bids on open ones. Won leases run on your infrastructure, usage settles hourly, and escrow pays out after the 24-hour dispute window.

Start with the provider docs Provider economics, explained Check network status

By operator type

Solutions for your kind of capacity