HIP-0024: Hanzo Sovereign L1 Chain Architecture

Abstract

Hanzo operates as a sovereign L1 chain on the Lux Network, meaning it maintains its own independent validator set rather than relying on primary network validation. This architecture provides Hanzo with full sovereignty over consensus parameters, validator economics, and network governance - essential for AI compute workloads requiring specialized infrastructure.

Motivation

Hanzo AI chose the L1 (sovereign) model for several strategic reasons:

AI compute requirements: Validators need GPU infrastructure for AI attestation
Full sovereignty: Control over consensus, upgrades, and economics
Custom validator requirements: TEE/SGX attestation for AI model integrity
Independent economics: AI-COIN staking separate from LUX
Specialized consensus: Tuned for AI inference workloads

Specification

Chain Configuration

chain:
  name: Hanzo
  type: L1
  chainId: 36963
  ticker: AI
  
validation:
  type: independent
  minValidators: 5
  stakingToken: AI
  
consensus:
  engine: nova
  parameters:
    k: 20
    alpha: 15
    betaVirtuous: 15
    betaRogue: 20

Genesis Configuration

{
  "config": {
    "chainId": 36963,
    "chainType": "L1",
    "validatorSet": {
      "type": "independent",
      "minValidators": 5,
      "stakingToken": "AI",
      "minStake": "2000000000000000000000"
    },
    "consensus": {
      "engine": "nova",
      "parameters": {
        "k": 20,
        "alpha": 15,
        "betaVirtuous": 15,
        "betaRogue": 20
      }
    }
  },
  "timestamp": "0x0",
  "gasLimit": "0xe4e1c0",
  "alloc": {}
}

Network Endpoints

Network	Chain ID	RPC Endpoint
Mainnet	36963	`https://api.hanzo.ai/ext/bc/hanzo/rpc`
Testnet	36962	`https://testnet.hanzo.ai/ext/bc/hanzotest/rpc`

Validator Requirements

Hanzo validators have specialized requirements beyond standard Lux validators:

validator:
  hardware:
    cpu: 32+ cores
    ram: 128GB+
    gpu: NVIDIA A100/H100 (for AI attestation)
    storage: 2TB+ NVMe
    
  software:
    tee: SGX/TDX enabled
    attestation: NVTrust compatible
    
  staking:
    minStake: 2000 AI
    lockPeriod: 14 days
    
  capabilities:
    - AI model inference
    - TEE attestation
    - FHE key share management

Validation Model

As an L1, Hanzo transactions are validated by its own validator set:

User Transaction → Hanzo Chain → Hanzo Validators → Consensus → Finality

Benefits:

Full control over validator requirements
Custom consensus tuning for AI workloads
Independent upgrade schedule
Specialized validator economics

Tradeoffs:

Must bootstrap and maintain validator set
Higher operational complexity
Independent security (not inherited)
Validator coordination for upgrades

AI Compute Integration

Hanzo validators perform AI-specific operations:

// AI Mining precompile (0x0300)
type AIMiningConfig struct {
    // ML-DSA signatures for quantum-safe attestation
    SignatureScheme string `json:"signatureScheme"` // "ML-DSA-65"
    
    // NVTrust verification for GPU compute
    AttestationType string `json:"attestationType"` // "NVTrust"
    
    // Reward calculation
    PrivacyLevels   []int  `json:"privacyLevels"`   // [1, 2, 3]
}

// Validator performs AI inference with attestation
func (v *Validator) ProcessAIRequest(req *AIRequest) (*AIResponse, error) {
    // 1. Verify TEE attestation
    if err := v.verifyTEE(req.Attestation); err != nil {
        return nil, err
    }
    
    // 2. Execute AI inference in enclave
    result := v.executeInEnclave(req.Model, req.Input)
    
    // 3. Sign with ML-DSA (post-quantum)
    signature := v.signMLDSA(result)
    
    return &AIResponse{Result: result, Signature: signature}, nil
}

Cross-Chain Communication

Hanzo uses Warp messaging for cross-chain operations:

// Send AI attestation to Zoo (L2)
IWarpMessenger(WARP_ADDRESS).sendMessage(
    ZOO_CHAIN_ID,
    abi.encode(modelHash, attestation)
);

// Receive compute request from C-Chain
function receiveComputeRequest(bytes calldata message) external {
    require(msg.sender == WARP_ADDRESS);
    // Queue AI compute job
}

Staking Economics

Hanzo has independent staking economics:

staking:
  token: AI
  minValidatorStake: 2000 AI
  minDelegatorStake: 25 AI
  rewardRate: 8% APY (base)
  aiMiningBonus: +2-5% (based on compute contribution)
  
slashing:
  downtime: 0.1% per hour (after 4 hour grace)
  doubleSigning: 5%
  invalidAttestation: 10%

Rationale

Why L1 over L2?

Factor	L2 (Primary Network)	L1 (Hanzo's Choice)
Validator control	Limited	Full
Custom requirements	Not possible	GPU/TEE required
Consensus tuning	Inherited	Customizable
Economics	LUX-based	AI-COIN based
AI integration	Limited	Native

Comparison with Zoo

Hanzo and Zoo represent the two chain models:

Chain	Type	Validators	Use Case
Zoo	L2	Primary network	Conservation, community AI
Hanzo	L1	Own set (GPU/TEE)	AI compute, sovereignty

This allows the ecosystem to demonstrate and optimize both approaches.

Security Considerations

Validator Security

Hanzo validators must maintain:

TEE attestation (SGX/TDX)
NVTrust GPU verification
ML-DSA key management
Secure enclave operations

Network Security

With 5+ validators and proper stake distribution:

Byzantine fault tolerance: 80%+
Economic security from AI staking
Slashing for malicious behavior

AI-Specific Security

Model integrity verification via attestation
Encrypted inference in TEE
Post-quantum signatures (ML-DSA)

Implementation

Deployment

# Create Hanzo chain (L1 mode)
lux chain create hanzo --type l1 --validators 5 --vm evm

# Configure validator requirements
lux chain config hanzo --validator-requirements gpu,tee

# Deploy to mainnet
lux chain deploy hanzo --mainnet

Validator Onboarding

# Register as Hanzo validator
lux validator join hanzo \
  --stake 2000 \
  --node-id NodeID-xxx \
  --tee-attestation attestation.json \
  --gpu-proof nvtrust.json

Hanzo Sovereign L1 Chain Architecture