Inspur AGX-2 AI Supercomputer

Inspur AGX-2 AI Supercomputer supports up to 8 GPUs in a 2U form factor with faster, more efficient computing performance for a wide range of demanding applications and environments.

Table 1 shows the quick spec.

Inspur AGX-2 AI Supercomputer provides these features:

Ultra-Density

Supports up to 8 GPUs in a 2U form factor with faster, more efficient computing performance for a wide range of demanding applications and environments.

Flexible Configuration

Supports 8NVIDIA Tesla P100 or 8NVIDIA Tesla P40 GPU.

10 GPU Topology configuration options for various Deep Learning or HPC applications.

Scale Up and Out

Supports GPU expansion box (JBOG) for up to 16 GPUs in one DP server system.

NVIDIA NVLink 2.0 ready for next generation NVIDIA Tesla GPU, up to 87.5% GPU interconnect bandwidth increase over the NVIDIA NVLink 1.0.

4*100G EDR InfiniBand port for cluster scale out

Green Design

Supports air cooling or air-liquid hybrid cooling, to provide lower PUE for datacenters.

Get More Information

Do you have any question about the AGX-2