FusionServer G560 Full-width Heterogeneous Computing Node

The Huawei FusionServer G560 is a full-width heterogeneous computing node that delivers high-density heterogeneous compute power and supports one-click topology switching. The server uses a fully modular design with independent CPUs and a heterogeneous unit. The FusionServer G560 is the best-in-class heterogeneous computing platform for AI training, High-Performance Computing (HPC), intelligent cloud, and database acceleration.

Superb performance, flexible configuration, and fully modular design

Exceptional Heterogeneous Computing Capabilities

  • A full-width heterogeneous computing node supports up to 8 FHFL heterogeneous accelerator cards
  • Support for GPUDirect RDMA and Peer-to-Peer interconnect technologies, enabling direct memory access for multiple GPU cards

Flexible Topology Configuration

  • Support of one-click topology switching and different CPU/GPU configuration ratios for diverse applications

Fully Modular Design

  • The design features both a decoupled CPU module and a heterogeneous module, supporting CPU and a heterogeneous unit for long-term evolution
  • Power Supply Units (PSUs), hard drives, and fan modules are designed in a fully modular style, supporting hot swap and redundancy backup, and simplifying O&M

Specifications

Product G560
Form Factor Full-width heterogeneous computing nodes
GPU Accelerator Card Up to 8 NVIDIA® Tesla® P100/P40
Processors 2 Intel® Xeon® E5-2600 v4 processors
Memory Up to 24 DDR4 DIMMs
Internal Storage Support 8 x 3.5-inch SAS/SATA HDDs, or 6 x 2.5-inch NVMe SSDs and 2 x 2.5-inch SSDs or SAS/SATA HDDs
RAID Support RAID 0, 1, 10, 5, 50, 6, or 60

Provides a supercapacitor to protect data in the case of power failures

Operating Systems Microsoft Windows Server, Red Hat Enterprise Linux, SUSE Linux Enterprise Server, Cent OS, and Ubuntu
Management On-board iBMC management module

Support management features such as IPMI, SOL, KVM over IP, and virtual media

Operating Temperature 5°C to 35°C

Algunas de nuestras REPRESENTACIONES