Dense AI Compute for Rack-Scale Workloads
The Dell PowerEdge XE9712 is built for demanding AI and HPC environments that require dense compute, high-bandwidth GPU communication, and rack-scale deployment. Based on NVIDIA Grace Blackwell Ultra GB300 architecture, it supports advanced AI model training, GenAI pipelines, and multi-rack GPU cluster designs.
Key Benefits and Features
- Combines NVIDIA Grace CPUs with Blackwell Ultra GPUs for accelerated AI compute.
- Uses NVLink CPU-GPU interconnect with 900GB/s coherent memory bandwidth.
- Supports direct liquid cooling and air cooling for high-density infrastructure.
- Fits into Dell IR9048 ORv3 rack infrastructure with 1RU XE9712 sleds.
- Includes embedded management options such as OpenBMC or iDRAC 10.
Ideal Use Cases
- AI model training and foundation model development.
- Generative AI infrastructure for enterprise or hyperscale environments.
- High-performance computing clusters requiring dense GPU acceleration.
- Rack-scale AI deployments using Dell IR9048 ORv3 infrastructure.
- Multi-rack GPU environments with high-speed east-west traffic requirements.
Technical Specifications
- Brand: Dell Technologies
- Model: PowerEdge XE9712
- Processor: Two NVIDIA Grace processors with 72 cores per processor
- Memory: 480 GB LPDDR5 ECC memory per CPU
- GPU: 4 Blackwell Ultra GPUs
- GPU Memory: 288 GB HBM3e per GPU
- Interconnect: 900GB/s coherent memory through NVLink CPU-GPU interconnect
- Drive Bays: Up to 8 x EDSFF E1.S hot-swappable NVMe drives
- Form Factor: 1RU XE9712 sled for IR9048 rack infrastructure
- Networking: 1G integrated LOM and 4 x16 CX8 OSFP ports
Build Dense AI Infrastructure with Dell XE9712
Choose the Dell PowerEdge XE9712 for rack-scale AI compute designed around Grace Blackwell Ultra architecture, NVLink connectivity, dense GPU acceleration, and Dell IR9048 rack integration.
