AI Training Server (NVIDIA DGX A100)*
Location: Online
Team: BCO
Process: AI; AI Training; GPU Computing; HPC
*Equipment of CAiRS
Specification
- A high-performance platform supporting AI training, inference, simulation, data processing, and other GPU-accelerated workloads
- 8 NVIDIA A100 GPUs, 40GB each
- 2 AMD EPYC 7742 processors
- 1TB RAM
- Each approved user will receive an isolated Ubuntu 24.04 LTS container with 2TB storage and 20GB MIG GPU allocation
Notes:
- The first booking requires a face-to-face meeting with a staff member to review the project scope, timeline, workload, and system requirements.
- Subsequent bookings may be approved through a simplified update process, provided the project requirements remain unchanged.
- Additional GPU and storage resources may be allocated according to project needs and platform availability.
