#USA #San Jose #NVIDIA Blackwell #ASUS AI POD #NVIDIA GB300

ASUS Unveils Revolutionary AI POD at GTC 2025

ASUS has taken a bold step into the future of artificial intelligence by introducing its latest ASUS AI POD, powered by the cutting-edge NVIDIA GB300 NVL72 platform. The launch at the GTC 2025, where ASUS is proudly participating as a diamond sponsor, signifies the company’s commitment to leading the AI infrastructure solutions sector.

A Game-Changer in AI Solutions

The ASUS AI POD is designed to empower businesses by providing robust solutions for extensive AI workloads. With game-changing specifications, the new server architecture integrates NVIDIA's Blackwell Ultra technology, ensuring enhanced computational efficiency and versatile capability to meet evolving data needs. This powerful technology enables businesses to confront substantial AI challenges head-on.

What sets this offering apart is its remarkable hardware configuration, which includes 72 NVIDIA Blackwell Ultra GPUs and 36 Grace CPUs, all housed within a convenient rack-scale framework. This design not only boosts AI processing capabilities but also offers up to 40TB of rapid memory per rack, ensuring that businesses can handle vast datasets with ease.

The ASUS AI POD supports advanced networking solutions, featuring integration with NVIDIA's Quantum-X800 InfiniBand and Spectrum-X Ethernet, reinforcing its position at the forefront of AI innovation. Designed with superior cooling mechanisms, the pod also supports groundbreaking workloads, facilitating trillion-parameter large language model (LLM) training and inference.

Extending Beyond AI POD

Alongside the AI POD, ASUS introduced a lineup of new servers optimized for AI applications. The ASUS XA NB3I-E12 features the NVIDIA B300 NVL16, offering enhanced AI performances. Other notable models include:

• - ASUS ESC NB8-E11 – outfitted with NVIDIA DGX B200's 8-GPU configuration.
• - ASUS ESC N8-E11V – featuring the NVIDIA HGX H200 architecture.
• - ASUS ESC8000A-E13P – designed for large-scale deployments, compatible with NVIDIA's MGX architecture.

These systems are meticulously crafted to ensure minimum latency in data processing while maximizing inference capabilities, critical for applications like AI reasoning and video inference.

The Supercomputing Leap with ASUS Ascent GX10

In a move that further cements its innovation credentials, ASUS also unveiled the ASUS Ascent GX10 AI supercomputer. This compact powerhouse, powered by the NVIDIA GB10 Grace Blackwell Superchip, delivers an astounding performance of 1,000 AI TOPS. It is tailored for demanding workloads, housing up to 200 billion parameter AI models, positioning it as an invaluable tool for AI developers and data scientists.

The Future of Edge AI

In addition to server-centric products, ASUS showcased its advanced Edge AI computing solutions. The PE2100N and PE8000G models are equipped with NVIDIA Jetson technology, enabling impressive performance for generative AI applications and real-time perceptions, expanding the possibilities for innovative solutions in industries like autonomous vehicles and intelligent video analytics.

Conclusion — ASUS Leading the Charge

With a strong foundation built on years of experience in AI infrastructure and a roster of forward-thinking products, ASUS is propelling itself as a critical player in AI innovation. The versatility and power of the new ASUS AI POD and its supporting systems empower businesses to optimize their AI strategies, allowing them to tackle upcoming challenges in data centers, across various industries.

As ASUS paves the path for future developments, it stands poised to redefine the scope of AI applications, enabling enterprises to achieve operational excellence and pioneering breakthroughs in technology.