Cloud AI Server Powered by AMD EPYC™
This 4U 2-socket AI server features 8x NVIDIA A100 GPUs with PCIe Gen4 interfaces and 2x AMD EPYC™ processors using the advanced 7nm process technology. The NF5468A5 delivers superior performance and a high-compatibility design to provide a powerful, reliable infrastructure for leading AI technologies and users worldwide.
![]() Leading Technology
|
![]() Consistent Quality
|
![]() Unparalleled Performance
|
![]() Comprehensive Ecosystem
|
Model | NF5468A5 |
---|---|
Processor | 2 x AMD EYPC™ processors interconnected via three xGMIs, with a TDP of 280 W |
GPU | 8 x A100, A40 or MI100 GPUs supporting PCIe interfaces, with a TDP of 350 W |
Memory | 32 x DDR4 RDIMMs at a transmission rate of 3200 MHz |
Rear PCIe | 2 PCIe 4.0 x16 and 1 OCP NIC 3.0 |
Storage | 24 x 2.5″ or 12 x 3.5″ SAS/SATA HDDs (up to 8 NVMe SSDs) and 2 M.2 SATA SSDs |
RAID Support | RAIDs 0, 1, 10, 5, 50, 6, and 60 in the optional configuration Cache protection with the supercapacitor RAID state transition and RAID configuration memory |
Operating System | Microsoft Windows Sever, Red Hat Enterprise Linux, Ubuntu Linux, CentOS, etc other mainstream operating systems |
System Heat Dissipation | System fans in N+1 redundancy |
Power Supply | 2+2 redundant 1600 W/2000 W/2200 W/ 3000W 80 Plus Platinum PSUs |
Operating Temp | 5 – 35°C / 41°F – 95°F |
Chassis | 478.8 mm (W) x 175.5 mm (H) x 830 mm (D) |
Weight in full configuration: | ≤70 kg |