Competitive Landscape

Several projects are already exploring different directions in the decentralized compute space. Below is a positioning comparison between the major players and DASP:

Bita's protocol competitiveness stems from three dimensions: verifiable delivery $\text{Verifiability}$, open accessibility $\text{Openness}$, and the network effect of provider scale $N_{\text{providers}}$. Platform-based competitors suffer from high vendor lock-in $\text{Vendor\_Lock}$, which limits their PCI.

Differentiation

DASP's core differentiator is not "more GPUs" but the protocol layer — we build the rules, not the data centers. From renting a single card to coordinating an entire network.

Protocol vs. Platform — Most competitors are platforms (aggregating supply, operating services); DASP is a protocol (defining the rules for matching, settlement, and verification that anyone can plug into). The protocol owns no data centers, operates no nodes, and locks in no users.

Verifiable Delivery — It is not just about renting out GPUs; on-chain receipts and verification mechanisms ensure that delivery quality is provable. Every compute delivery comes with an independently verifiable on-chain credential.

# Delivery Verification — On-chain Credential Generation
def verify_delivery(task_id: str, output: bytes) -> Receipt:
    output_hash = sha256(output)
    commitment = sha256(
        task_id.encode() + output_hash + timestamp_bytes()
    )
    merkle_leaf = MerkleTree.insert(commitment)

    receipt = Receipt(
        task_id=task_id,
        output_hash=output_hash.hex(),
        commitment=commitment.hex(),
        merkle_root=merkle_leaf.root.hex(),
        merkle_proof=merkle_leaf.proof,
    )
    tx = chain.submit_receipt(receipt)
    return receipt.with_tx(tx.hash)

AI-Native Design — The protocol was designed for AI workloads from day one: inference billing models, training job orchestration, and agent-based task scheduling — not simply slapping GPUs onto general-purpose compute.

Inference costs are billed by input/output token counts plus a base duration fee. $c_{\text{input}}, c_{\text{output}}$ are the unit prices for input and output respectively (per million tokens), and $c_{\text{base}}$ is the endpoint keep-alive fee.

Roadmap

# Phase 1 Testnet — Core Interface
dasp testnet deploy \
    --chain sepolia \
    --contracts escrow,registry,benchmark

dasp provider register \
    --gpu A100 --capacity 4 --stake 1000

dasp job submit \
    --spec spec.yaml --budget "100 tUSDC"

# Phase 2 — Reputation Query & Dispute Resolution
score = dasp.reputation.get("0x8B3f...a91D")
# → ReputationScore(
#     overall=0.94, uptime=0.997,
#     sla_compliance=0.96,
#     benchmark_percentile=92,
#     dispute_rate=0.008,
#     total_orders=1847
# )

dispute = dasp.dispute.open(
    order_id="ORD-2024-0917",
    evidence={
        "expected_hash": "sha256:a9f3c2...",
        "received_hash": "sha256:7b2e1f...",
    },
)
# dispute.status → "under_review"

# Phase 3 — Enterprise Private Supply Pool + Cross-chain
pool = dasp.PrivatePool(
    name="acme-corp-gpu",
    access="whitelist",
    members=["0xAcme...", "0xPartner..."],
    sla={
        "uptime": 0.999,
        "data_residency": "EU",
        "compliance": ["SOC2", "GDPR"],
    },
)

bridge = dasp.CrossChainBridge(
    source="ethereum", target="solana",
)
tx = bridge.settle(order_id="ORD-2024-1203")
# tx.source_hash → "0xe7a1..."
# tx.target_hash → "5Kj9p..."

Governance

Protocol parameters (fee rates, penalty thresholds, admission criteria) are adjusted through on-chain governance. Treasury fund usage requires a proposal and voting process. The goal of governance is to let the protocol evolve collectively by its participants — demand-side users, providers, and validators — rather than being dictated by a single team.

Voting weight uses the square root of the staked amount to curb whale concentration, and applies a logarithmic bonus for longer lock-up periods to encourage long-term participation.

A proposal must satisfy both conditions: the proportion of approval-vote weight exceeds the threshold $q = 66.7\%$, and the total voting weight exceeds the minimum quorum $Q_{\min}$ (15% of total staked amount).

# Governance Proposal — Adjust Protocol Parameters
proposal = dasp.governance.propose(
    title="BIP-042: Adjust protocol fee 2.5% → 2.0%",
    params={
        "protocol_fee": 0.02,
        "burn_rate": 0.40,
        "validator_share": 0.35,
    },
    discussion_period=7 * 86400,   # 7-day discussion period
    voting_period=3 * 86400,       # 3-day voting period
    quorum=0.15,                   # 15% staked amount quorum
)
# proposal.status → "discussion"
# proposal.vote_url → "https://gov.bita.forum/BIP-042"

Network Economics

Network value is positively correlated with the number of providers $N_p$, the number of demand-side users $N_d$, the average transaction volume $\bar{T}$, and the average reputation $\mathcal{R}_{\text{avg}}$. The reputation term constrains the value contribution of low-quality providers through exponential decay.

The rate of change in token supply is determined by the interplay of new emissions $R_{\text{emission}}$, protocol fee burns $B$, and staking lockups $L$. As network activity increases, $B(t)$ grows, creating deflationary pressure.

// Token Economics Model
{
  "total_supply": "1,000,000,000 DASP",
  "allocation": {
    "ecosystem_incentives": "40%",
    "team_&_advisors":      "15%",
    "treasury":             "20%",
    "public_sale":          "15%",
    "liquidity":            "10%"
  },
  "fee_structure": {
    "protocol_fee": "2.0%",
    "burn_rate": "30% of fees",
    "validator_share": "50% of fees",
    "treasury_share": "20% of fees"
  }
}