VVDN Model ConsoleL40S inference control
Connecting to AI server
Inference orchestration

Models, ready when
your applications are.

Load the exact runtime you need, release VRAM when you are done, or switch the complete server configuration with one click.

Control accessConfirm credentials to enable model changes
Read only

Everything is read-only until confirmation. Password is never stored.

—models loaded
2NVIDIA L40S
92GB total VRAM

Application presets

Unloads conflicting runtimes and starts every service the application needs.

What is loaded on each GPU

Live model placement, memory, utilization, and temperature from the inference server.

All locally available models

Managed runtimes plus every model found by the bounded hourly inventory scan.