EnterpriseAI Platform
Glossary
| Term | Description |
|---|---|
| HAMi | GPU virtualization and scheduling foundation. |
| vGPU | Virtualized GPU resource unit. |
| Memory isolation | Limits the GPU memory scope available to a workload. |
| Oversubscription | Allows requested resources to exceed physical capacity under policy control. |
| Tenant | Logical boundary for resources and permissions in the platform. |