Stay Cool, Scale Fast
Supercharging AI Factories for Enterprise

ASUS AI POD with NVIDIA HGX™ Rubin NVL8

The ASUS XR R1I12B-M1BHFS is a hybrid-cooled, rack-scale AI Factory built on the NVIDIA HGX™ Rubin NVL8 platform — pre-validated with 9 compute trays, 72 NVIDIA Rubin GPUs, and 18 Intel® Xeon® 6776P processors. Purpose-built for enterprises scaling generative AI, agentic AI, and large-scale inference, it delivers the AI capability, operational efficiency, and deployment speed to build trusted AI Factories with confidence.
ASUS AI POD with NVIDIA HGX Rubin NVL8 in an enterprise AI factory data center, with server racks extending into a mountain landscape.
ASUS AI POD with NVIDIA HGX Rubin NVL8 in an enterprise AI factory data center, with server racks extending into a mountain landscape.
ASUS 3U compute tray viewed from the front and top, showing the compact rackmount chassis and front I/O.
1 2 3 4

1Air-Cooled CPU Sled
Simplified Design, Faster deployed time

2Flexible Storage Options
Bigger space for U.2/E1.S

3Fully Cable-less system
Board-to-Board Design, High Computing Density

4Up to 50% Less Maintenance Time
Faster Servicing, Simplified Stock & Less Downtime

Abstract blue AI computing visualization representing accelerated AI workloads.

Faster Time-to-Train

Built to shorten every training cycle
  • Massive parallel GPU scaling
  • Ultra-low latency interconnect
  • Sustained performance under full load
Enterprise professional working with an AI computing system, representing enterprise AI deployment.

Balanced Performance & Efficiency

Engineered to do more with less
  • Low-latency processing
  • High throughput
  • Cost-efficient operation
Enterprise data center server racks illustrating scalable AI infrastructure deployment.

Deploy with Confidence

Expertise that carries through the full lifecycle
  • Faster time-to-deployment
  • Consistent uptime and reliability
  • Simplified maintenance and flexibility

NVIDIA HGX™ Rubin NVL8,
Powering Every AI and HPC Workload

The NVIDIA HGX™ Rubin NVL8 platform integrates eight NVIDIA Rubin GPUs with sixth-generation NVLink® interconnects, delivering up to 10x more token factory throughput and matching training performance with 4x fewer GPUs versus HGX B200. Paired with dual Intel® Xeon® 6776P processors and a fully modular, tool-less design, the ASUS compute tray translates this platform leap into lower cost per token, faster time-to-model, and a stronger return on every enterprise's AI Factory investment.

Boost Token Factory Throughput

10x
Token Factory Throughput
VS NVIDIA HGX B200*
4X Fewer GPUs
To Train Same Model
VS NVIDIA HGX B200*

*Source: NVIDIA. Projected performance, subject to change. Token throughput based on Kimi K2-Thinking (FTL≤500ms, ISL=4K, OSL=4K); HGX Rubin NVL8 with Sparse NVFP4 vs. HGX B200 with Dense NVFP4. Training comparison based on DeepSeek-R1 pretrained on 15T tokens, 4K sequence length.

Accelerating the Next Generation of Agentic AI

NVIDIA HGX Rubin NVL8 compute platform with multiple GPU modules and high-bandwidth interconnect architecture.
8
NVIDIA Rubin GPUs
2.3
TB of HBM4 memory
3.6 TB/s
NVLink between GPUs via NVLink Switch chip

A 3U Compute Tray Built for Flexibility and Serviceability

Each 3U compute tray pairs 8 NVIDIA Rubin GPUs with dual Intel® Xeon® 6776P processors in a hybrid-cooled design — direct liquid cooling for the GPU sled and air cooling for the CPU sled — delivering a more easy-maintained thermal solution than full liquid cooling. This fully cable-less, board-to-board architecture and simplifies assembly, while generous storage space give customers flexible storage options for U.2/E1.S configurations. The fully air-cooled CPU sled also cuts maintenance time by up to 50%, striking the ideal balance of performance and operational simplicity so enterprises can scale AI Factories with fewer components and less downtime.
  • System architecture diagram of the ASUS 3U compute tray showing CPU, GPU, memory, storage and connectivity components. System architecture diagram of the ASUS 3U compute tray showing CPU, GPU, memory, storage and connectivity components.
  • Internal view of the ASUS 3U compute tray showing the internal board layout, cooling components and system modules.

Leading the Era of Super-Intelligent AI Reasoning

XR R1I12B-M1BHFS Key Features

ASUS AI POD with NVIDIA HGX™ Rubin NVL8

ASUS AI POD system specification panel showing rack-scale configuration and key system specifications.

North-South Networking

North-South Networking

Power Shelves

Management Node

File System

Object Storage

Management switch 

Power Shelves

9 x Compute Trays
• 18 x Intel® Xeon® 6 processors
• 72 x NVIDIA Rubin GPUs

Power Shelves

ASUS AI POD - 576p GPUs
Specification items
1 Rack
1 SU (=8 Racks)
CPU support
18 x Intel Xeon 6776P Processors
144 x Intel Xeon 6776P Processors
GPU support
72 x GPUs (= 9x NVIDIA Rubin NVL8)
576 x GPUs (= 72x NVIDIA Rubin NVL8)
Compute Tray
9 x NA NR1I-E12L(R)
72 x NA NR1I-E12L(R)
Switch (N/S)
9 x Nvidia BlueField-4 DPU
72 x Nvidia BlueField-4 DPU
Switch (E/W)
1 x Nvidia NVSwitch SN2201
9 x Nvidia NVSwitch SN2201

ASUS Professional Services

Your Trusted Partner from Factory to Future

ASUS provides end-to-end support throughout your entire lifecycle, from rigorous L10/L11 factory validation and aging tests to meticulous air-cushion transport and vibration-resistant packaging. Leveraging NVIDIA DSX's unified platform for optimizing tokens per watt, ASUS helps companies build large-scale AI Factories. We integrate DSX-ready systems, managing the full deployment lifecycle from assessment to scalable inference, to reduce risks and accelerate real-world implementation.
  • logo of deployment

    Testing and Validation

  • logo of shipment

    Packaging and Delivering

  • logo of datacenter

    Deployment and Management

ASUS Professional Services for AI infrastructure deployment, shown with multiple enterprise data center racks and infrastructure equipment.

FAQs

  • Why does XR R1I12B-M1BHFS  use a 3U hybrid-cooling design instead of full liquid cooling like the 2U XR R1I12B-M1BLFS ?
    Why does XR R1I12B-M1BHFS  use a 3U hybrid-cooling design instead of full liquid cooling like the 2U XR R1I12B-M1BHFS ? XA NR1I-E12L is built for data centers that need to run HGX Rubin NVL8 GPU compute but aren't yet ready for full direct liquid cooling (DLC) across every component. It applies direct liquid cooling to the GPU sled only, while the CPU sled uses air cooling (6 x system fans). This hybrid architecture, housed in a 3U (447mm x 800mm x 131mm) chassis with 54V busbar power, lets customers adopt Rubin NVL8 GPU density while reducing the facility-side liquid-cooling infrastructure required at initial deployment.