How was the base rate chosen?
The metering equation is mathematically consistent. But why 0.001 DIU per GPU-second instead of 0.0007 or 0.0018? The base rate is derived from four factors — and it is adjustable, not fixed forever.
Modeled Economics — Not Yet Observed
The economic framework on this page is structurally sound and calibrated against published cloud GPU pricing, industrial electricity costs, and standard treasury mechanics. However, the specific figures are projections, not measurements. The following metrics require real customer operations to move from modeled to observed:
Real payout history across active contributors
Actual spend vs. equivalent hyperscaler inference
GPU-hours consumed ÷ GPU-hours available, measured live
Redemption behavior during demand shocks
Contributor and enterprise retention over time
What is defensible today
The framework — metering equation, floor-target-ceiling band, unified reserve model, prepay issuance — is the structurally sound part. It is designed to produce correct incentives regardless of the specific numbers.
What is not yet demonstrated
Whether the modeled rates produce sustainable contributor retention, real enterprise savings, and stable reserve behavior under stress. These are empirical questions answered by adoption — not by protocol design.
As real customers onboard, each metric above transitions from modeled to observed and is published with its source (modeled projection vs. production telemetry).
The question
The metering equation determines how much Vital is issued per unit of compute. But the equation has a base rate — a constant that anchors the output. Why is it 0.001 DIU per GPU-second? Is it based on electricity costs? GPU depreciation? Enterprise willingness to pay? Target contributor earnings? The answer is all four.
Four factors, one rate
Electricity cost
Sets the floorThe marginal cost of powering a GPU for one second. This is the floor — a contributor should not earn less than the electricity they spend. The base rate is anchored to average industrial electricity pricing in the primary operating regions.
GPU depreciation
Sets the floorThe hardware cost amortized over useful life. A GPU that costs $10,000 and runs for 3 years has a depreciation cost per second. The base rate must cover this so contributors are not subsidizing hardware wear with free compute.
Enterprise willingness to pay
Sets the ceilingWhat enterprises currently pay for equivalent cloud inference (per GPU-second on AWS, Azure, GCP). The base rate must be below this ceiling — otherwise enterprises have no reason to switch from hyperscaler pricing.
Target contributor earnings
Sets the targetEnough to make participation worthwhile after electricity and depreciation. If the rate is too low, contributors leave. If it is too high, enterprises leave. The rate sits in the band where both sides benefit.
The balance
The base rate sits in a band between the floor and the ceiling:
Electricity + depreciation. Below this, contributors lose money on every second.
Contributor earnings that make participation worthwhile — enough margin above the floor to justify the effort, risk, and hardware commitment.
Enterprise willingness to pay (hyperscaler equivalent pricing). Above this, enterprises have no reason to switch.
The current base rate (0.001 DIU/GPU-second) is set within this band. It is low enough that enterprises pay less than equivalent cloud inference, and high enough that contributors cover electricity, depreciation, and earn a margin.
The rate is adjustable
The base rate is not fixed forever. It is a parameter that governance can adjust based on real market feedback:
The honest answer
The current base rate is a starting estimate, calibrated against published cloud GPU pricing and average industrial electricity costs. It has not yet been tuned against months of real utilization data. As the platform processes real workloads, the rate will be adjusted based on actual contributor retention, enterprise conversion, and utilization patterns. The framework (four factors, floor-target-ceiling band) is the defensible part; the specific number is a starting point that will be refined with operating data.
DOSFI Sovereignty — Closed Core / Open Surface
DOSFI Core is sovereign and closed. All educational content on this page is generated exclusively from open-surface materials — SDKs, ABIs, documentation, tutorials, examples, and conceptual explanations. Internal details of the DOSFI Runtime, Scheduler, Router, Readiness Engine, Privacy Model, Power/Thermal Model, Vital internals, DIU internals, MNE internals, and Mesh-Native App runtime internals are not revealed, described, or inferred.
No proprietary algorithms, system code, or architectural diagrams capable of enabling replication of the DOSFI Core are published. When internal logic is referenced, only high-level conceptual explanations are provided. DOSFI's sovereignty, mission-lock, and non-replicable architecture are preserved at all times.
See the full metering equation
The six-factor equation that determines issuance, with the base rate as its anchor.
View end-to-end transaction