Provider Policy

Terms, requirements, and guidelines for GPU providers on WorldLand Cloud.

Eligibility Requirements

Hardware Requirements

Component	Minimum	Notes
GPU	NVIDIA GTX 1080 or higher	CUDA compute capability 6.0+
VRAM	8GB	Higher VRAM enables more workloads
System RAM	16GB	Per GPU recommended
Storage	100GB SSD	NVMe preferred
Network	100 Mbps symmetric	1 Gbps recommended
Public IP	Required	Static IP strongly recommended

Software Requirements

• Ubuntu 20.04 LTS or 22.04 LTS
• Latest NVIDIA drivers (525+)
• Docker with NVIDIA Container Toolkit
• Kubernetes components (kubelet, kubeadm)

Network Requirements

• Port forwarding capability
• NodePort range accessible (30000-32767)
• Stable internet connection
• Low latency to major regions

Service Level Agreement (SLA)

Uptime Requirements

Tier	Minimum Uptime	Penalty
Standard	95%	Warning
Enhanced	99%	Required for premium listing
Enterprise	99.9%	Priority matching

Calculating Uptime

Uptime % = (Total Time - Downtime) / Total Time × 100

Downtime Definition

Downtime includes:

• Node unreachable
• GPU unavailable
• Network connectivity issues
• Unplanned maintenance

Planned Maintenance

• Notify 24 hours in advance (via dashboard)
• Maximum 4 hours planned maintenance per month
• Coordinate during low-usage periods

Pricing Guidelines

Setting Your Price

You have full control over your pricing:

GPU Model	Market Range (WL/hour)	Recommended
RTX 4090	0.40 - 0.70	0.50
RTX 3090	0.30 - 0.50	0.35
RTX 3080	0.25 - 0.40	0.30
Tesla T4	0.20 - 0.35	0.25
A100 40GB	0.80 - 1.50	1.00
A100 80GB	1.20 - 2.00	1.50

Price Changes

• Prices can be updated anytime
• Changes apply to new jobs only
• Existing jobs retain original price

Quality Standards

Container Requirements

All customer containers must have:

• ✅ Full GPU access (nvidia.com/gpu)
• ✅ Root SSH access
• ✅ Specified CPU/Memory allocation
• ✅ Network connectivity
• ✅ CUDA drivers functional

Performance Standards

Metric	Requirement
Container startup	< 60 seconds
SSH availability	< 30 seconds after Running status
GPU initialization	nvidia-smi responsive
Network latency	< 100ms to container

Monitoring

WorldLand monitors:

• Node health via heartbeat
• GPU availability
• Resource utilization
• Customer feedback

Prohibited Activities

Providers Must NOT:

❌ Terminate customer containers without cause
❌ Access customer container data
❌ Throttle resources below allocation
❌ Misrepresent hardware specifications
❌ Share customer information
❌ Run unauthorized workloads on customer containers

Immediate Termination

The following result in immediate removal:

• Security breach of customer containers
• Fraudulent resource reporting
• Repeated SLA violations
• Illegal content or activities

Revenue & Payments

Revenue Split

Customer Payment
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Provider: 90% │██████████████████░░│
Protocol: 10% │██░░░░░░░░░░░░░░░░░░│
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Payment Schedule

• Payments processed on-chain
• Real-time settlement upon job completion
• Automatic deposit to provider wallet

Payment Disputes

1. Customer initiates dispute
2. Evidence review (logs, metrics)
3. Resolution within 48 hours
4. Arbitration if unresolved

Data Protection

Provider Responsibilities

• No access to customer container contents
• No logging of customer data
• Secure deletion of container storage after job termination
• Network isolation between customer containers

Ephemeral Storage

All customer data is ephemeral:

• Deleted immediately on job termination
• No backups or snapshots by default
• Customers responsible for their own data backup

Compliance

Required Compliance

Providers must comply with:

• Local laws and regulations
• Export control requirements
• GDPR (if serving EU customers)
• Network security best practices

Audit Rights

WorldLand reserves the right to:

• Request hardware verification
• Audit resource availability
• Review operational practices
• Conduct security assessments

Account Management

Suspension

Provider accounts may be suspended for:

• SLA violations
• Customer complaints
• Security concerns
• Policy violations

Appeal Process

1. Receive suspension notice
2. Submit appeal within 7 days
3. Review by WorldLand team
4. Decision within 14 days

Voluntary Withdrawal

To stop providing resources:

1. Set status to "Maintenance"
2. Wait for active jobs to complete
3. Request removal from network
4. Settle any pending payments

Updates to Policy

This policy may be updated periodically. Providers will be notified of material changes:

• 30 days advance notice for significant changes
• Immediate effect for security-related updates
• Continued operation implies acceptance

Questions?

For policy questions, contact support@worldland.cloud or visit our Discord.

Next Steps

• How to Provide - Setup guide
• Provider Rewards - Earnings structure