Solana Fee Model Simulation Maps Out Who Pays More Under SGP0003

Solana is preparing to overhaul how transaction fees are calculated, and a new simulation suggests the change would redistribute costs sharply across applications while materially increasing SOL burned. Analyst @MostlyData_ modeled the impact of governance proposal SGP0003, introduced on Aug. 3, 2026 alongside SIMD0553. The proposal would replace Solana's flat 5,000 lamport base fee with a two-part structure: a fixed 2,500 lamport inclusion fee paid to the block leader, plus a variable resource fee starting at 0.1 lamports per requested compute unit (CU) that is fully burned. Simulation results: over-requested compute becomes expensive The model's central takeaway is that many transactions request more compute than they ultimately use. On average, transactions ask for about 20% more CUs than they consume. Under the current flat-fee system, that padding has no additional cost; under SGP0003, it directly raises fees. Routers and aggregators appear most exposed. The simulation estimates average fee increases of roughly 0.000068 SOL for Jupiter, 0.00010 SOL for Titan, and 0.00012 SOL for DFlow. It also finds that only about 28% of transactions would keep fee increases below 10% if resource pricing aligns tightly with actual consumption. Validators could see relief on vote costs The proposal may benefit validators in one specific area: optimized vote transactions could be about 12.3% cheaper under the new fee model, according to the simulation. Burn rate could jump by multiples Solana currently burns about 648 SOL per day from signature fees. The simulation projects SGP0003 could lift daily burns to a range of 1,500 to 9,000 SOL. The proposal was introduced alongside a separate measure to double Solana's disinflation rate from 15% to 30%, combining higher burn with a faster slowdown in new token issuance. Winners, losers, and a "tax on inefficiency" SGP0003 effectively prices inefficiency: applications that tightly calibrate compute requests could keep fees near the old flat rate, or potentially lower. Apps that routinely over-request CUs would pay a premium for the added buffer. Central limit order book (CLOB) market makers may face particularly meaningful cost math. Their high transaction volume and heavier compute requirements mean even small per-transaction increases can compound into material expenses. The simulation flags CLOB market making as especially exposed. For DeFi, routers such as Jupiter already account for a large share of Solana transaction flow. Even modest per-swap increases could be passed on to users or squeeze aggregator margins.