The network measures total solving compute in AMAFLOPS — like FLOPS, but not standard floating-point ops, since every solve attempt includes a blake3 step. It targets 100 peta-AMAFLOPS and pays out that percentage of the solver emission each epoch. The rest accrues — it is not handed to a small set of solvers.
Paid per computeYou earn your compute's share of the 100 peta-AMAFLOPS target — a fixed rate per op while participation is 10–100%.
No dilutionRun 1 GPU, add 100 more → 101× the ops, 101× the emission. New solvers never shrink your rewards below the 100 target.
10% floorAt least 10% always pays out. Below the floor, fewer solvers split it — each op earns more.
2,000,000 solve attempts/s × 25,722,880 ops = 0.0514 peta-AMAFLOPS of the 100 target
Card solving at 2M attempts/s (RTX 5090 class)
Raw reward
per epoch (~13.9 h)
≈ 58.0 AMA
per day
≈ 100.2 AMA
Raw protocol reward before pool fee. Above the 100 target the budget caps and extra solving power dilutes; emission decays ≈0.3%/day.