Skip to content

Ethereum vs Solana, by cluster

Every topic page on this site ends with an “Ethereum vs Solana” section (§8). This page pulls the single most defining comparison out of each of those 62 sections into one table per cluster, so you can scan the whole site’s chain-by-chain differences at once. Every cell is copied or tightly paraphrased from its source page’s §8 — no new figures are introduced here. Where a page states there is no equivalent on one chain, the cell reads “n/a — <reason>” exactly as that page states it.

TopicEthereumSolanaPage
Execution modelSequential per block; state discovered at runtimeParallel across non-conflicting write sets; accounts declared per instruction/foundations/accounts-and-execution/
What a token isA contract implementing an interface (ERC-20/721/1155/4626/6909); arbitrary code per tokenA mint account processed by one shared Token/Token-2022 program; behavior toggled via declared extensions/foundations/token-standards/
Fee market designDynamic base fee (±12.5%/block, burned) plus a tip to the builder; separate exponential blob feeFixed 5,000-lamport base fee plus a priority fee per compute unit, priced locally per contended account/foundations/fee-markets/
Block-building pipelinePublic mempool → searcher → builder → relay → proposer (MEV-Boost)No mempool; transactions go straight to the scheduled leader, with Jito auctioning bundles at the top of the block/foundations/mev-supply-chain/
Scaling thesisRollups plus an L1 that specializes in data availability and settlementA single chain scaled vertically — bigger compute budgets, faster consensus/foundations/roadmaps/
TopicEthereumSolanaPage
Canonical CFMM and swap costUniswap v2 (x·y=k, 30bp fee), v3/v4 ranges; typical swap $5–$50 (as of 2026-04)Raydium AMM v4/CPMM, Orca and Raydium CLMM, Meteora DLMM; typical swap $0.001–$0.01 (as of 2026-04)/exchange/cfmm-math/
Arbitrageur’s share of LVR at a 30bp fee (model)≈12%, driven by ~12s block time≈2.5%, driven by ~0.4s slots/exchange/impermanent-loss-vs-lvr/
Concentrated liquidity: discretisation and dominant useTicks (1.0001i1.0001^i); blue-chip and stable pairs, since gas makes narrow ranges costly to rebalanceTicks (Orca, Raydium) or discrete bins (Meteora DLMM); the venue of choice for new launches and memecoins/exchange/concentrated-liquidity/
Canonical stable-asset AMMCurve StableSwap / StableSwap-NG (up to 8 tokens); Orbital proposed for nn stablecoinsSaber (a Curve-style port), Meteora DAMM stable pools, and 0.01% CLMM tiers/exchange/stable-pools/
Live embedded TWAMMFraxswap (since 2022-06); an example hook exists in Uniswap v4-peripheryn/a as an embedded TWAMM — Jupiter Recurring (DCA) and Meteora DLMM limit orders substitute/exchange/twamm/
Live uniform-price batch protocolCoW Protocol (~1% of DEX volume, as of 2024-07)n/a — per-order Dutch auctions (Drift JIT) and off-chain RFQ (JupiterZ) instead/exchange/batch-auctions/
Live spot order books and quote costClober on L2s; a quote update costs roughly 40k–50k gas at L1 pricesPhoenix, OpenBook v2; 100k–300k compute units (fractions of a cent) but must win priority/exchange/order-books/
Aggregator dominanceDiversified: CoW 29.2%, 1inch 24.8%, Kyber 12.1% of $31.09B (2025-07, Helius citing The Block)Jupiter alone routes 86.4% of $64.1B in aggregator volume (2025-07, Helius citing Blockworks)/exchange/routing-and-aggregation/
Flagship intent systemsUniswapX, CoW Protocol, 1inch Fusion, AcrossJupiterZ RFQ (default since 2024-12), Drift JIT auctions/exchange/rfq-and-intents/
Pool programmabilityUniswap v4 hooks with fixed callbacks; am-AMM and LVR-hook designs mostly in researchNo hook framework; proprietary AMMs re-centre via cheap oracle updates, over 60% of SOL/USDC volume (2025-07)/exchange/v4-hooks-and-am-amm/
Dominant on-chain price sourceUniswap v2/v3 TWAP (on-chain accumulator)Pyth’s pull oracle with an explicit confidence interval/exchange/amm-oracles/
Where toxic order flow is measuredPublic mempool plus on-chain history; JIT liquidity is ~0.3% of Uniswap v3 volume (as of 2022-07)No public mempool; toxicity is managed privately inside closed-source proprietary AMMs/exchange/lp-toxicity-and-jit/
Production private AMMNone — but the deepest published research on the impossibility (Angeris, Evans, Chitra)None — Token-2022 Confidential Transfers solve a narrower problem (hiding transfers, not swaps)/exchange/zk-dex-limits/
Dominant prediction-market mechanismCentral limit order book (Polymarket, on Polygon)Prop AMMs (Jupiter Forecast, launched 2026-06)/exchange/prediction-market-amms/
Dominant token-launch primitiveGDA/VRGDA for NFTs and targeted schedules; creator or protocol funds an initial poolPump.fun’s constant-product virtual-reserve curve; fully synthetic at creation, zero creator capital required/exchange/bonding-curves-and-launches/
TopicEthereumSolanaPage
Flagship pooled marketsAave V3, CompoundKamino Lend, marginfi/P0, Save/lending/money-markets/
Dominant rate modelTwo-slope kinked curve (Aave/Compound lineage); fixed rates via a separate market (Yield Protocol, Morpho Midnight)Multi-point interpolated curve (up to 7 points), configured per bank/reserve; no fixed-rate discovery market documented/lending/interest-rate-models/
Oracle-uncertainty handling in health checksNot addressed by a confidence interval in this page’s Aave sourcesmarginfi explicitly prices assets low and liabilities high inside a capped 5% Pyth confidence band/lending/collateral-ltv-health/
Worst documented liquidation failureMakerDAO Black Thursday (2020-03-12) — zero-bid auctions under gas-price congestionSolend whale crisis (2022-06-19) — governance forced to choose disorderly liquidation or an emergency takeover/lending/liquidations/
Isolated-market primitiveMorpho Blue (5 immutable parameters), Euler v2’s EVK vaultsKamino’s “Create Market” — an isolated market with custom risk parameters/lending/modular-lending/
Best-documented flash-loan incidentbZx, February 2020 — oracle and liquidation-check manipulation funded by a flash loanmarginfi, September 2025 — a flash-loan account-tracking bug, caught and patched before any exploit/lending/flash-loans/
Risk-engine designReserve-by-reserve parameters, some updatable by a permissioned Risk Admin without a full governance votemarginfi’s deterministic, Just-In-Time engine reading live oracle and bank state in the same instruction/lending/risk-engines/
TopicEthereumSolanaPage
Dominant fiat-backed coinUSDC, deeply integrated across DeFiUSDC, over 70% of stablecoin supply by share (as of 2025-05, Helius)/stablecoins/fiat-backed-and-cctp/
Flagship CDP stablecoinDAI/USDS (Sky), crvUSD (Curve)n/a at comparable scale — Solana’s dollar liquidity instead comes from fiat-backed issuers and delta-neutral designs/stablecoins/cdp-stablecoins/
Where the delta-neutral hedge livesEthena’s hedges sit on centralized derivatives exchanges, anchored to Ethereum-based collateraln/a — Solana’s own attempt (UXD Protocol, fully on-chain) shut down in August 2024 for lack of product-market fit/stablecoins/delta-neutral/
Largest algorithmic collapseIron Finance / TITAN (June 2021, ~$2B TVL wiped); TerraUSD/LUNA ran on its own Cosmos chain but shaped Ethereum design debatesn/a — no comparable Solana-native algorithmic stablecoin in these sources/stablecoins/algorithmic-failures/
Documented de-peg incidentUSDC/SVB, March 2023 — reserve illiquidity at the issuer, propagated into DAI via its PSMCashio, March 2022 — a smart-contract exploit, unrelated to reserves/stablecoins/pegs-and-psm/
Purpose-built payment chainTempo — an independent EVM L1, sub-second finality, stablecoin-denominated feesn/a as a dedicated payments chain — Solana Pay standardizes point-of-sale payments on mainnet itself/stablecoins/payments/
TopicEthereumSolanaPage
Dominant perpetuals venue architectureOrder book (dYdX, historically) and oracle-priced pool (GMX)Hybrid JIT-auction + order book + AMM (Drift); pure oracle-priced pool (Jupiter Perps/JLP)/derivatives/perpetual-futures/
Live power perpetualsOpyn Squeeth — live 2022-2024, since shut downn/a — none found; Solana’s derivatives ecosystem trades only linear (p=1p=1) perpetuals/derivatives/power-perps-and-everlasting-options/
Pooled-writer optionsHegic (Arbitrum), Opyn, and the AMM-native PanopticPsyOptions / PsyFi/derivatives/options/
Dominant yield-tokenization protocol and scalePendle — roughly $1.04bn TVL across 12 chains (2026-07), having peaked near $8.9bn (2025-08)Exponent — emerging, with no primary TVL report or AMM whitepaper found in sources/derivatives/yield-tokenization/
Dominant live prediction-market venuePolymarket (on Polygon), using Gnosis’s CTF and UMA’s optimistic oracleMetaDAO — a live futarchy/decision-market protocol offering “Futarchy as a Service”/derivatives/prediction-markets/
TopicEthereumSolanaPage
Issuance ruley=cF/Dy = cF/\sqrt{D}, no ceilingFixed disinflation schedule: 8% initial → -15%/yr → 1.5% terminal floor/staking/issuance-and-staking-economics/
Liquid-staking market concentration~70% of the LST market sits in one token (stETH), ~31% of all staked ETHFragmented — Jito ~35%, Marinade ~42% of a small LST segment; under 5% of staked SOL is in LST form at all/staking/liquid-staking-tokens/
Primary restaking protocols and slashing statusEigenLayer, Symbiotic — slashing is live, redesigned around isolated “unique stake” per operator setJito (Re)staking — slashing is explicitly “in current development,” not live/staking/restaking/
Real LRT stress eventThe ezETH depeg, April 2024 (~79% drop in under an hour, $56M+ liquidated)None recorded in these sources — no comparable liquid restaking token market exists yet/staking/lrt-risks/
TopicEthereumSolanaPage
Dominant toxic MEV channelPublic-mempool sandwiches (declining) and builder-level extractionJito bundles and private validator mempools/mev/mev-taxonomy/
Who builds the block, and the escrow layerSpecialized builders under MEV-Boost; the proposer signs blind, trusting a relay to escrow the blockThe scheduled leader builds and executes its own block; no relay — Jito is an off-chain auctioneer for bundles only/mev/pbs-and-epbs/
Who captures the block-auction proceeds todayThe proposer; burn and execution-ticket designs are proposed but unshippedThe leader and stakers, via Jito tips (a 6% Jito fee)/mev/mev-burn-tickets-focil-timing/
Default order-flow-auction user refund90% of backrun bid value (MEV-Share, MEV Blocker)Usually none — protection (DontFront) rather than a rebate/mev/order-flow-auctions/
Production encrypted-ordering approachTEE builders (BuilderNet ~40% of blocks, Unichain 100%, as of 2025-07); Shutter’s threshold encryption live on Gnosis ChainBAM TEE nodes, in a permissioned rollout (as of 2025-07)/mev/encrypted-mempools/
Pending-transaction visibilityPublic mempool, though private orderflow now exceeds half of all gas (2025)None in protocol — the Jito engine and private validator mempools see flow instead/mev/solana-mev/
Live fast-confirmation / based-sequencing examplesTaiko (based rollup; preconfirmations shipped August 2025); Unichain (TEE builder; 200ms flashblocks, August 2025)n/a as rollups — Solana scales one state machine; Alpenglow targets ~150ms finality instead of a preconf layer/mev/based-rollups-and-preconfs/
TopicEthereumSolanaPage
Dominant bridge trust modelGuardian networks (Wormhole) and native burn-and-mint (CCTP) both prominentThe same two models — Solana is a first-class chain in both/cross-chain/bridge-designs/
Total documented bridge-hack losses in this set (as of 2024-01)≈$1.05B across Nomad, Ronin, Harmony, Multichain and Orbit combined$326M (Wormhole) — the exploited program lived on Solana, but backing collateral sat on Ethereum/cross-chain/bridge-hacks/
Underlying interop problemDozens of independent rollups, each with separate gas, bridges and finalityn/a — a single execution environment has no rollup-fragmentation problem to solve/cross-chain/cross-l2-interop/
Dominant fast-bridge fulfillment modelEscrowed and live today (Across); a bonded upgrade is proposed (Across Prime)deBridge-style validator settlement; no bonded model documented/cross-chain/intent-solvers/
TopicEthereumSolanaPage
Dominant price-oracle patternChainlink push (Data Feeds), Uniswap on-chain TWAPPyth pull (Pythnet aggregation), consumed by most DeFi protocols/oracles/push-vs-pull-vs-twap/
Documented major oracle-manipulation incidents in this setbZx, Harvest, Inverse Finance, Venus (4 of 5)Mango Markets (1 of 5, but the largest by dollar amount)/oracles/oracle-manipulation/
Oracle-free lendingBlend (Blur) — live, NFT-focused, auction-based solvency with no price computedn/a — not documented; Solana’s major money markets rely on Pyth instead/oracles/oracle-free-designs/
TopicEthereumSolanaPage
Dominant bug familyUnsafe external calls (reentrancy) and identity/indirection bugs in composed contractsMissing or incomplete account validation (signer, owner, type, program identity)/security/bug-classes/
Rows in the incident table and largest single loss34 rows; Bybit, $1.46B–$1.5B (2025-02)18 rows; Wormhole, $326M (2022-02, exploited program lived on Solana)/security/incident-timeline/
Dominant governance/admin-key incidentBeanstalk — a flash-loan governance attack, $77M, purely mechanical, no human deceptionDrift Protocol — socially-engineered pre-signed governance transactions, $285M, a human-boundary attack/security/trusted-hardware-and-threat-models/
TopicEthereumSolanaPage
Base-layer governanceDeliberately minimized; changes move through social consensus and hard forksValidator stake-weighted SIMD feature-gate votes; no tokenholder DAO over the base protocol/governance/governance-and-tokenomics/
TopicEthereumSolanaPage
Tokenized-equity model breadthRobinhood tracker tokens under MiFID II on an Arbitrum L2, for EU usersFour distinct live models simultaneously: xStocks (tracker), Backpack (convertible), Ondo GM (KYC’d interface), Opening Bell (native shares)/tradfi/tradfi-convergence-and-rwas/
TopicEthereumSolanaPage
What is hiddenDepositor/withdrawer identity (deposit-withdrawal unlinkability); amounts often still exposed via fixed-denomination poolsTransfer amount and balance; sender and recipient addresses stay public/privacy/privacy-and-compliance/

Read across all 62 tables above, six threads keep recurring.

Execution model. Everything else in this comparison is downstream of one fork in 2015-2020-era design: Ethereum discovers what a transaction touches by running it, so it must execute sequentially; Solana requires transactions to declare their accounts up front, so non-conflicting transactions run in parallel (/foundations/accounts-and-execution/). That single choice explains why Ethereum’s answer to congestion is rollups (spin up more execution environments) while Solana’s is vertical scaling (make the one environment faster) — the entire roadmaps comparison collapses to this point (/foundations/roadmaps/).

Fee markets. Ethereum prices blockspace once, globally, with a base fee that moves slowly and is burned; Solana prices it per-account, with a fixed base fee and a priority fee that only escalates where there is actual contention (/foundations/fee-markets/). This is why a popular NFT mint can wreck Ethereum’s entire fee market for hours (the Otherside mint, $150M+ in failed transactions) while an equivalent Solana event mostly prices out only the contended account.

MEV supply chain. Ethereum built an explicit, layered market for block-building because its public mempool made ordering visible and valuable: searcher → builder → relay → proposer, with MEV-Boost formalizing what used to be chaotic gas-price bidding (/foundations/mev-supply-chain/, /mev/pbs-and-epbs/). Solana never had a mempool to make visible, so its MEV market is about latency and access instead — who can send a transaction to the leader fastest, and who can pay Jito’s tip auction for a guaranteed bundle position (/mev/solana-mev/). Tellingly, both ecosystems are now converging on the same fix from opposite directions: TEE-attested block building (BuilderNet and Unichain on Ethereum, BAM on Solana) (/mev/encrypted-mempools/).

Liquidity venues. Ethereum’s slow, expensive blocks pushed liquidity toward AMMs that only need to be right once every ~12 seconds, and toward off-chain price discovery (RFQ, intents) to avoid paying that cost repeatedly (/exchange/cfmm-math/, /exchange/rfq-and-intents/). Solana’s cheap, fast slots instead pushed liquidity toward actively quoted proprietary AMMs that re-price dozens of times per second — passive xy=kx\cdot y=k liquidity increasingly serves only long-tail launches there (/exchange/v4-hooks-and-am-amm/). Both chains ended up dominated by one aggregator layer (Jupiter on Solana at 86.4% share; a more fragmented mix on Ethereum), which is where most retail price competition actually happens now (/exchange/routing-and-aggregation/).

Oracles. Ethereum standardized on push feeds (Chainlink) and its own AMMs’ time-weighted averages, because a deep on-chain pool already existed to average; Solana standardized almost entirely on one pull oracle (Pyth), because no Solana AMM produced an equivalently battle-tested on-chain history mechanism (/oracles/push-vs-pull-vs-twap/, /exchange/amm-oracles/). The threat models differ accordingly: Ethereum’s oracle incidents are almost all “move a thin pool for one block” (bZx, Harvest); Solana’s one major incident (Mango Markets) was self-referential — the attacker was simultaneously the price-setter and the price-taker for his own collateral (/oracles/oracle-manipulation/).

Stablecoin rails. Ethereum’s stablecoin ecosystem is the most structurally diverse — fiat-backed (USDC), CDP-backed (DAI/USDS, crvUSD), delta-neutral (USDe) and, historically, algorithmic (the failed designs) all coexist at real scale (/stablecoins/cdp-stablecoins/, /stablecoins/algorithmic-failures/). Solana’s dollar liquidity is comparatively narrow — overwhelmingly fiat-backed USDC, with a growing but not dominant delta-neutral presence, and no CDP or algorithmic stablecoin at systemic scale — but its speed and near-zero fees make it the more natural settlement rail for payments rather than pure DeFi collateral, which is exactly why Solana Pay exists on general-purpose mainnet while Ethereum’s answer was to build an entirely separate payments-first chain, Tempo (/stablecoins/payments/).

The pattern across all six threads is the same: the two chains rarely disagree on what to build — an AMM, an oracle, a block-building market, a stablecoin — they disagree on where to put the resulting cost, because one chain’s blocks are slow and expensive and the other’s are fast and cheap.