Superior Heterogeneous Computing Capability
أ 4 U chassis supports a maximum of eight full-height full-length dual-slot heterogeneous accelerator cards with a maximum power consumption of 350 W or 32 half-height half-length heterogeneous accelerator cards with a maximum power consumption of 75 W.Supports GPUDirect RDMA and Peer-to-Peer.
Flexible Heterogeneous Topology Configuration
Supports one full-width or two half-width heterogeneous computing nodes, one-click topology switching, and multiple topologies with CPU/GPU configuration ratios of 1:2, 1:4, و 1:8.
Fully Modular Design
Adopts decoupled CPU modules and heterogeneous modules to support the long-term evolution of CPUs and heterogeneous components.The PSUs, محركات, and fan modules are modular and support hot swap and redundancy.
Main Application
Scenarios
|
نموذج |
Node/Processors |
Accelerator Card |
DIMM
|
تخزين |
AI inference, AI render, HPC,
image/video analysis,
and database
|
FusionServer G5500 V6 |
2 x 3rd Generation Intel® زيون® القابلة للتطوير معالجات |
8 x DW GPUs |
32(8 تيرابايت)
|
24×2.5ʺ /3.5 ʺ |
FusionServer G5500 V7 |
2 x 4th or 5th Gen Intel® زيون® القابلة للتطوير معالجات |
10 x DW GPUs |
32(8 تيرابايت)
|
24*2.5ʺ /3.5ʺ SAS/SATAʺ; |
|
AI inference, AI render,
and voice/image/video
analysis
|
FusionServer G5200 V7 |
2 x 4th or 5th Gen Intel® زيون® القابلة للتطوير معالجات |
10*SW GPUs; |
32(8 تيرابايت)
|
32*2.5 ʺ /3.5ʺ SAS/SATA +4* NVMe; |
AI inference, HPC, و
image/data analysis
|
FusionServer G8600 V7 |
2 x 4th or 5th Gen Intel® زيون® القابلة للتطوير معالجات |
8-GPU Modules |
32(8 تيرابايت)
|
8*2.5ʺ SAS/SATA+8*NVMe |
—
|
FusionServer G5500 V5 chassis |
يدعم 1 x full-width node; support 2 x half-width nodes |
— |
—
|
— |
AI inference and render
|
FusionServer G530 V5 half-width node |
Two 2nd Generation Intel® زيون® القابلة للتطوير معالجات |
16 x SW GPUs and 4 x DW GPUs |
24(3 تيرابايت)
|
4 x 3.5ʺ + 2 x 2.5ʺ; 2 x 2.5ʺ; |
AI training, HPC, و
render
|
FusionServer G560 V5 full-width node |
Two 2nd Generation Intel® زيون® القابلة للتطوير معالجات |
8 x DW GPUs |
24(3 تيرابايت)
|
8 x 3.5ʺ; 8 x 2.5ʺ; |