Three phases, one record per asset
Deployment and baseline
When a GPU enters service we record its baseline: initial thermal profile, power characteristics, and the workload the borrower described in the loan application.
Initial state
Factory spec and expected use
Workload profile
Projected usage pattern
Risk score
Starting risk assessment
Ongoing monitoring
Measured performance is checked against the baseline continuously. An alert fires when an asset drifts from the assumptions the loan terms rest on.
Utilization trend
Measured vs. projected usage
Thermal health
Operating temperature trend
Stress events
Spikes, throttling, errors
Depreciation and salvage value
Each asset gets its own depreciation curve from measured wear and performance decay. The salvage value projection updates as the asset ages.
Performance decay
Measured degradation
Economic life
Updated projection
Salvage value
Estimate from measured wear
What you get
Per-asset reporting
- →Full performance history for every GPU
- →Measured vs. expected utilization
- →Thermal stress, quantified
- →Maintenance event log with the effect on life
Portfolio analytics
- →Portfolio health score
- →Borrower performance ranking
- →Salvage value forecast
- →Risk concentration