NVIDIA GB200 NVL72

ASUS ESC AI POD

Explore more AI breakthroughs in a single rack.

Unimaginable AI Unleashed

ESC NM2N721-E1 with NVIDIA GB200 NVL72

Extensive ASUS expertise in crafting AI servers with unparalleled performance and efficiency has been bolstered by its collaboration with NVIDIA. One of the highlights of the showcase is the NVIDIA Blackwell-powered scale-up larger-form-factor system – ESC AI POD with NVIDIA GB200 NVL72. The full rack solution is a symphony of GPUs, CPUs and switches harmonizing in lightning-fast, direct communication, turbocharging trillion-parameter LLM training and real-time inference. It’s equipped with the latest NVIDIA GB200 Grace Blackwell Superchip and fifth-generation NVIDIA NVLink technology, and supports both liquid-to-air and liquid-to-liquid cooling solutions to unleash optimal AI computing performance.

KEY FEATURES

Unparalleled AI performance: All-new ASUS ESC AI POD with the NVIDIA® GB200 NVL72 system and the NVIDIA GB200 Grace Blackwell Superchip
Full ASUS AI server lineup: From hybrid servers to edge-computing deployments, ready for training, inference, data analytics and HPC
Software-defined data center solutions: End-to-end services tailored to enterprise needs, from top-notch hardware to comprehensive software
A liquid cooled, rack scale solution that boasts 36 Grace CPUs and 72s Blackwell GPU
5 ^th Generation NVIDIA NVLink technology within single domain
NVIDIA BlueField ® 3 to enable cloud networking and composable storage

5 X

LLM Inference

^{vs. NVIDIA H100 Tensor Core GPU}

1 X

LLM Training

^{vs. H100}

12 X

Energy Efficiency

^{vs. H100}

1 X

Data Processing

^{vs. CPU}

Unlocking Real-Time Trillion-Parameter Models

Rack-Scale Architecture for Real-TIme Trillion-Parameter Inference and Training

ASUS AI Pod connects 36 Grace CPUs and 72 Blackwell GPUs in a rack-scale design. Liquid-cooled, rack-scale solution that boasts a 72-GPU NVLink domain that acts as a single massive GPU and delivers 30X faster real-time trillion-parameter LLM inference.

The GB200 Grace Blackwell Superchip is a key component of the NVIDIA GB200 NVL72, connecting two high-performance NVIDIA Blackwell Tensor Core GPUs and an NVIDIA Grace CPU using the NVIDIA® NVLink®-C2C interconnect to the two Blackwell GPUs.

GB200 NVL72 Specifications

	GB200 NVL72	GB200 Grace Blackwell Superchip
Configuration	36 Grace CPU : 72 Blackwell GPUs	1 Grace CPU : 2 Blackwell GPU
FP4 Tensor Core²	1,440 PFLOPS	40 PFLOPS
FP8/FP6 Tensor Core²	720 PFLOPS	20 PFLOPS
INT8 Tensor Core²	720 POPS	20 POPS
FP16/BF16 Tensor Core²	360 PFLOPS	10 PFLOPS
TF32 Tensor Core	180 PFLOPS	5 PFLOPS
FP32	6,480 TFLOPS	180 TFLOPS
FP64	3,240 TFLOPS	90 TFLOPS
FP64 Tensor Core	3,240 TFLOPS	90 TFLOPS
GPU Memory \| Bandwidth	Up to 13.5 TB HBM3e \| 576 TB/s	Up to 384 GB HBM3e \| 16 TB/s
NVLink Bandwidth	130TB/s	3.6TB/s
CPU Core Count	2,592 Arm® Neoverse V2 cores	72 Arm Neoverse V2 cores
CPU Memory \| Bandwidth	Up to 17 TB LPDDR5X \| Up to 18.4 TB/s	Up to 480GB LPDDR5X \| Up to 512 GB/s

Technological Breakthroughs

Blackwell Architecture

The NVIDIA Blackwell architecture delivers groundbreaking advancements in accelerated computing, powering a new era of computing with unparalleled performance, efficiency, and scale.

Learn More >

NVIDIA Grace CPU

The NVIDIA Grace CPU is a breakthrough processor designed for modern data centers running AI, cloud, and HPC applications. It provides outstanding performance and memory bandwidth with 2X the energy efficiency of today’s leading server processors.

Learn More >

Fifth-Generation NVIDIA NVLink

Unlocking the full potential of exascale computing and trillion-parameter AI models requires swift, seamless communication between every GPU in a server cluster. The fifth-generation of NVLink is a scale–up interconnect that unleashes accelerated performance for trillion- and multi-trillion-parameter AI models.

Learn More >

NVIDIA Networking

The data center’s network plays a crucial role in driving AI advancements and performance, serving as the backbone for distributed AI model training and generative AI performance. NVIDIA Quantum-X800 InfiniBand, NVIDIA Spectrum™-X800 Ethernet, and NVIDIA BlueField®-3 DPUs enable efficient scalability across hundreds and thousands of Blackwell GPUs for optimal application performance.

Learn More >