IO + Ensurable Systems: Cardano Maintenance Initiative — Independent Cost Analysis
Cardano Treasury Withdrawal #100 (62.13M ADA / 9 months)
This proposal (Treasury Withdrawal #100) passed the DRep + CC voting threshold and was ratified on-chain at epoch 633 on the morning of 2026-05-25 (will auto-enact at ep634). The analysis below was written pre-vote and is preserved as reference on cost reasonableness. Five other IO bundle proposals ratified in the same epoch (~131.5M ADA total = 37.6% of the 350M NCL), while Pogun, Blockfrost, and IO+Midgard L2 expired without reaching the ratification threshold.
62.1M ADA (= 9 months, ~$20.7M annualized @ $0.25/ADA) decomposed into 3 buckets aligned to peer chain primary client team scope:
Core observation: "Cardano is too expensive vs peers" is NOT empirically supported, and "simply reasonable" is also not the case. Bucket A + B (~$11.5M, 55.7%) is comparable to peer client team responsibility, Bucket C (~$9M, 43.4%) is the cost of Cardano-specific architectural / design choices. Per-FTE rate is $230K/yr, lower than peer range ($250-400K) — labor cost ratio is efficient. The evaluation core is the reasonableness of Bucket C scope (DevOps incl bootstrap/testnet ops, Haskell ecosystem coordination, multi-component integration, Cardano Blueprint implementation-independent spec).
📐 Methodology
Decomposes Treasury Withdrawal #100's activity-level breakdown (12 categories) into 3 buckets, comparing each to peer chain primary client team scope. This is single-client-team apples-to-apples comparison, NOT ecosystem-wide aggregated total.
Comparison set: Ethereum Geth (EF subset), Solana Agave (Anza), Bitcoin Core (aggregated grants), partial Tezos Trilitech (closest formal-spec chain peer).
Sources: Proposal PDF + official reply (2026-05-11), EF 2024 Annual Report (verified), public GitHub data (Anza, Geth team size), independent cardano-git-analytics methodology.
✅ Bucket A: Pure Client Maintenance
$5.5M(26.5% of total)Peer comparison: Directly comparable to peer chain primary client maintenance: Geth (EF subset) $3-7M, Anza maintenance-only ~$5-9M, Bitcoin Core $5-12M aggregated.
Verdict: Within range. The pure client maintenance scope is comparable to peer chain teams.
🟡 Bucket B: Broader Client Team Responsibility
+$6M(+29.2%, cumulative w/ Bucket A: $11.5M)Bucket A + the 3 activities below = peer chain broader engineering team responsibility. Each compared to peer chain similar work to calculate reasonableness.
Performance & Tracing
$2.3M(11.1%)✅ within-rangeGeth/Anza performance teams: included in core engineering. Estimated $1-3M/yr per chain.
Systematic ledger performance + node optimization (block production, diffusion, validation) + benchmarking for releases and hard forks. Critical for Praos timing requirements + DoS attack prevention.
Tech Debt reduction
$2.88M(13.9%)✅ within-rangeIndustry standard ~5-15% of engineering budget. Ethereum Geth team: continuous refactoring within feature work, hard to isolate. Anza: ~10-15% (estimated).
Cardano-specific complexity: Conway era ledger transitions, Haskell ecosystem migrations (GHC version bumps), Plutus version transitions. IO claims 20% tech debt reduction in 2025 enabled Leios delivery in 2026.
Documentation (Cardano Blueprint)
$0.87M(4.2%)✅ within-rangeEthereum Yellow Paper: ~$0.3-0.5M/yr maintenance. EIP Editor team (EF): ~$0.5-1M/yr. Bitcoin: BIP editors volunteer-based, $0. Cardano Blueprint is broader scope (consensus + network + ledger + Plutus).
Implementation-independent comprehensive spec. Enables alternative node implementations (Amaru / Dingo / Gerolamo) to verify correctness. Materially larger scope than EF's split Yellow Paper + EIPs.
→ Bucket B total $6.05M is comparable to slightly lower than peer chain broader engineering team responsibility. Calculation transparency: each activity has peer chain equivalent at similar cost order of magnitude.
🚨 Bucket C: Cardano-Specific Scope
+$9M(+43.4%, total $20.7M)Scope derived from Cardano-specific architectural / design choices. No direct peer equivalent, or peer chains structure differently. Each component broken down with relevant peer ecosystem references.
DevOps incl bootstrap relay + testnet ops + disaster recovery
$5.18M(25%)🟡 comparable but high- Ethereum Foundation: runs Sepolia + Holesky testnets + EF devops infrastructure. Estimated $2-5M/yr embedded in Foundation operations.
- Solana Foundation: Tour de SOL testnet + bootnodes + monitoring. Devops costs embedded in $5-15M/yr Foundation operations.
- Bitcoin: testnet3/signet/regtest community-run, minimal cost (~$0).
Cardano includes CI/CD system maintenance, Haskell compiler tooling coordination, disaster recovery per CIP-135, bootstrap relay operations (= will be reduced ~$200K/yr after Genesis deployment), testnet maintenance (Preview + Preprod).
$5.18M is within Eth ($2-5M) and Solana ($5-15M Foundation) range. 25% of single proposal is high BUT this reflects Cardano's architectural choice of in-protocol bootstrap relays + active testnet pair.
Downstream dependency management (Haskell ecosystem)
$2.88M(13.9%)⚠️ high (architectural)- Ethereum Go-Ethereum: Go toolchain is Google-maintained, stable releases. Dependency coordination overhead minimal (estimated <$0.5M/yr embedded in core team).
- Solana Agave: Rust toolchain is corporate-maintained (Rust Foundation), stable. Dependency overhead low (<$0.5M/yr).
- Filecoin (similar formal-spec chain): Rust + Go, similar low overhead.
- Tezos OCaml: more comparable — Trilitech/Nomadic Labs maintain OCaml-specific tooling. Cost not publicly disclosed but estimated $0.5-1.5M/yr.
Haskell ecosystem is materially smaller than Go/Rust. IOG historically responsible for keeping Cardano build green across GHC versions. Plus libsodium custom integration, Plutus-specific Haskell library coordination. This is a structural cost of Cardano's Haskell-first architecture.
$2.88M (13.9%) is notably higher than Go/Rust chain equivalents (~$0.5M). Tezos OCaml ($0.5-1.5M) is the closest peer. Cardano sits above this range, reflecting GHC niche ecosystem + custom build infrastructure.
Multi-component maintenance (Plutus Core + DB-Sync + CLI integration)
$2.00M(~10-12%%)🔵 Cardano architecture- Ethereum: EVM is bundled in Geth (= integrated, no extra cost). Smart contract VM cost embedded.
- Solana: sealevel parallel runtime + accountsdb bundled in Agave (= integrated).
- Bitcoin: Script (= minimal VM) bundled in Bitcoin Core.
- Cardano Plutus Core: SEPARATE codebase from cardano-node (independent GitHub repo with different commit authors confirmed = structurally distinct from Geth's EVM bundling).
Cardano architectural choice: Plutus Core (smart contract VM) is a separate Haskell codebase. DB-Sync (indexer) is also separate. Cardano CLI is separate. This adds maintenance integration overhead between components — versioning coordination, conformance testing across components, release synchronization. Estimated 5-10 FTE total across these adjacent codebases.
$2M is the Cardano-architectural cost of split codebases vs Eth/Solana's bundled approach. No direct peer reference because peer chains chose bundled architecture. This is a design tradeoff cost, not necessarily inefficient.
Component Maintenance — Haskell-specific libs (libsodium, technical debt, guardrails, CC identity scripts)
$0.97M(~4-5%%)✅ within-range- Ethereum: libsecp256k1 maintenance shared with Bitcoin (community + EF + Blockstream).
- Solana: cryptographic libraries maintained as Rust crates, minimal coordination overhead.
- Bitcoin Core: libsecp256k1 + custom crypto, $0.5-1M/yr (Chaincode + Brink + Spiral grants).
Cardano uses libsodium (Cardano-specific version), plus Conway era guardrails, plus CC identity scripts. Smaller than Bitcoin's libsecp256k1 maintenance burden.
$0.97M is within Bitcoin Core's $0.5-1M range for cryptographic library maintenance.
Cross-checking 4 Bucket C components against peer ecosystem references shows each cost order of magnitude is justifiable: (1) DevOps within Eth/Solana Foundation devops range, (2) Haskell coordination comparable to slightly higher than Tezos OCaml (closest peer), (3) Multi-component has no peer reference because peer chains chose bundled architecture, (4) Component maintenance (libsodium etc.) within Bitcoin Core range. **Bucket C cost itself is NOT ad-hoc inflated; it is the structural cost of Cardano architectural design choices**. The evaluation question should refocus from 'is cost reasonable' to 'are these design choices themselves the right trade-off for ecosystem outcomes'.
- Cardano Treasury Withdrawal #100 proposal PDF (IPFS: QmYKV7vbX9pwNxst2qWpmtM2k6LRJB4NKKNFJSQyh7zojz)
- Official IO reply to social media discussion (2026-05-11)
- Ethereum Foundation 2024 Annual Report (ethereum.foundation/report-2024.pdf, L1 R&D figures verified exactly)
- Anza / Solana Labs public engineer count (LinkedIn / news coverage 2024-2025)
- Bitcoin Core funding entities (Spiral, Chaincode Labs, MIT DCI, Brink — public grants)
- Independent cardano-git-analytics methodology (12-month contributor analysis)
- Tezos Foundation operations references (CHF-denominated, partial disclosure)
This analysis is independent community work and does not represent the views of IO, Ensurable Systems, or any official organization. Figures cross-verified to the extent of publicly available sources.