Skip to Content
AEX-2UA2

SuperEdge 2U · AI inference / training server

AEX-2UA2

A flexible rear-access 2U GPU server for on-premises LLMs and enterprise agentic AI: two double-width GPUs in a 420 mm short-depth chassis.

2U
420 mm short depth
2 × GPU
double-width
2000 W
1+1 redundant Titanium
30 kg
weight

Highlights

AEX-2UA2

  • Powered by Intel® Xeon® 6 processors with high performance cores.
  • Dual double-width GPUs for seamless generative AI, LLMs, and deep learning.
  • Flexible GPU support, including NVIDIA L40S, NVIDIA RTX PRO™ 6000 Blackwell Server Edition, NVIDIA H200 NVL, and NVIDIA RTX PRO™ Blackwell GPUs.
  • Compact 2U short-depth (420mm) rear-access rackmount server.
  • High-bandwidth architecture for fast, efficient data transfer in demanding enterprise tasks.

Specifications

AEX-2UA2

AEX-2UA2
ChassisForm FactorRear Access 2U Rackmount
Dimensions (W x D x H)438mm x 420mm x 88mm
ProcessorSingle Intel® Xeon® 6700-series with P-cores
GPUMax CountUp to 2 Double-width PCIe Passive GPUs or 1 Double-width PCIe Active GPUs (blower type)
QualifiedL40S, RTX PRO™ 6000 Blackwell Server Edition, H200NVL with NVLink bridge, Others*; *Please contact us
System Memory8 x Channel Memory, DDR5 6400MT/s with ECC
StorageDrive BaysFront: 4x Hot-swappable E1.S, up to PCIe Gen5, up to 15mm E1.S
M.22 x M.2, up to PCIe Gen5 x4, M-Key 2280/22110
RAID SupportNVMe RAID 0/1/5/10 support (Intel® VROC RAID key required)
Input/Output1 x RJ45 1Gbe Dedicated BMC LAN port; 1 x MiniDP port; 2 x USB 2.0 ports
Front PanelButtonsSystem Power; UID/BMC reset
LEDPower status; UID; Storage Activity; Network Activity; Information
PCIe Expansion slots2 x PCIe Gen5 x16 link (PCIe x16 slot, FHFL) for 2 Double-width GPUs with NVLink bridge; 1 x PCIe Gen5 x8 link (PCIe x16 slot, FHFL)*; 1 x PCIe Gen5 x16 link (PCIe x16 slot, FHHL); *Slot not avaliable when 2 Double-width AOC is installed.
PowerPower SupplyDefault: 2 x 2,000W Redundant (1+1) Titanium Level (96%) power supplies
Power InputDefault 2,000W power supply:; 1,000W: 100-127V~, 12A, 47-63Hz; 2,000W: 220-240V~, 10A, 47-63Hz
Power ConnectorsIEC 60320 C13
Package (W x D x H)662mm x 642mm x 305mm
Weight30kg
MISC. FunctionTrusted Platform Module (TPM) 2.0; Intel® Platform Firmware Resilience (Intel® PFR)
TemperatureOperating Temperature: 10°C ~ 35°C (50°F ~ 95°F); Non-operating Temperature: -30°C to 60°C (-22°F to 140°F); *10°C ~ 30°C for RTX PRO 6000 server edition configurations
HumidityOperating Relative Humidity: 8% to 80% (non-condensing); Non-operating Relative Humidity: 8% to 90% (non-condensing)
OS SupportLinux Ubuntu Server; Microsoft Windows Server 2019 and 2022
CertificationCE、FCC、UL、KCC

Ordering Information

AEX-2UA22U Rackmount short-depth GPU server with 2000W 1+1 Redundant Titanium level PSU

Application

Which Sector, Which System?

Edge AI Applications

A short-depth 2U server with dual double-width GPUs for generative AI, LLMs and deep learning; with NVIDIA H200 NVL, RTX PRO 6000 Blackwell Server Edition or L40S it runs models in-house.

Sector page

Data Centre

A 420 mm short-depth rear-access 2U rackmount server: dual double-width GPUs and a high-bandwidth architecture.

Sector page

N4060-VSFX

A compact PCIe card with the NVIDIA GeForce RTX 4060

Product card

IA380E-QUFL

A low-profile, single-slot Intel Arc A380E PCIe card

Product card

AIP-FR68S

A GPU-scalable edge AI workstation for on-prem AI, LLM and advanced vision inference

Product card

All Aetina Products

Why Björn Tech?

Integration in a Single System

We plan the Aetina product together with the enclosure, thermal solution and Innodisk storage / memory as one project.

The Right Choice for the Application

We choose the right model with you based on your power, space, temperature and I/O budget.

A Local Point of Contact

As the authorised representative in Türkiye we are your single point of contact for quotes, product selection and project support.

Talk to our engineering team about your project.

Contact Us