EcoTensor AIEcoTensor AIDecision Intelligence
Simulation mode

Synthetic telemetry. No physical GPU control is enabled.

Mission ControlGPU ClusterWorkloadsDecision CenterImpactProof & BaselineAudit LogSettings
Sense → Score → Decide → Explain
Human review remains in control.
Live simulator2s refresh
API offlineBalanced mode

Policy and simulation

Settings

Control the recommendation policy without changing the existing decision engine or physical infrastructure.

Operating mode

Safety boundaries override every policy mode.

Simulation environment

The frontend displays the current prototype boundary clearly.

Prototype
Telemetry source
Synthetic GPU telemetry
Refresh interval
2 seconds
Decision persistence
JSONL
Physical GPU control
Disabled
The interface never claims to control real hardware. Production telemetry, authentication, infrastructure integrations, and durable multi-user storage remain roadmap work.

API readiness

Browser-to-Python adapter status

Offline
Start FastAPI on port 8000 to manage settings.

Safety hierarchy

How policy and risk interact

01

Thermal safety

High or critical thermal risk can force workload movement.

02

Deadline safety

Unsafe deadline risk blocks optional delay.

03

Workload safety

Real-time and high-priority jobs receive stronger protection.

04

Mode preference

Performance, Balanced, or Eco applies only after safety checks.