Home / AI DC
Future direction
Private AI. Closer to your data.
A future edition of the platform, built around GPU workloads. Edge DC and the first pilot come first.
A complete AI stack
The proposed AI edition brings GPU compute, fast storage, high-speed networking and cooling designed for higher density into one modular unit.
- Local model hosting and AI inference
- Private AI services and model storage
- Kubernetes and workload scheduling
- GPU health, capacity and power monitoring
Sensitive data can stay close
Government, banks, universities, research teams and industrial companies may need AI near their data. Access controls, storage policies and external service dependencies must be designed for each workload.
Start with the workload
An optional GPU workload in the pilot can help establish memory, storage, networking and cooling needs. No GPU configuration, training capacity or performance target is confirmed yet.
SMARTBOX × CONNECTORZZZ
Start small. Prove it in the real world.
Choose a real workload. Build one unit. Measure what matters.