#Taiwan #Taipei #ASUS #AI Supercomputer #ExpertCenter

ASUS Launches the ExpertCenter Pro ET900N G3: A Groundbreaking AI Supercomputer

On June 23, 2026, ASUS unveiled the ExpertCenter Pro ET900N G3, a revolutionary deskside AI supercomputer engineered to bring advanced data-center-class AI performance directly into the hands of enterprises, AI developers, and data scientists. By harnessing the power of the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip and built on the NVIDIA DGX Station GB300 architecture, this cutting-edge machine is set to redefine the capabilities of AI computation.

Designed for Modern AI Workloads

As the demand for sophisticated local AI infrastructure continues to rise, the ExpertCenter Pro ET900N G3 addresses these challenges by delivering unparalleled performance for large-scale AI training and inference. The system supports a range of advanced workloads, including fine-tuning large language models (LLM), generative AI tasks, deep learning research, and the deployment of autonomous AI agents. Equipped with an impressive coherent memory of 748GB, it allows developers to train larger models while minimizing workflow bottlenecks typically associated with conventional workstations.

Unprecedented AI Performance

With a staggering capacity of up to 20 petaflops (PFLOPS) of AI performance, the ET900N G3 is built to handle demanding computational tasks right from your desktop. This means businesses can now execute complex AI workloads locally with improved control, lower latency, and enhanced data privacy. The shift towards local AI development ensures that organizations can rely less on cloud infrastructures, reducing operational costs and maintaining a competitive edge in innovation.

Seamless Integration into Work Environments

The design of the ET900N G3 integrates work-friendly features suited for office environments without compromising performance. By combining advanced AI capabilities with enterprise-grade scalability, organizations can deploy this supercomputer without needing a specialized data center. It supports NVIDIA's comprehensive AI software stack, providing a streamlined environment for machine learning and analytics, thus helping to foster seamless computational performance across interconnected AI workflows.

Enhancing Enterprise AI Execution

With the growing significance of agentic AI, the ExpertCenter Pro ET900N G3 ensures secure local AI execution within organizational infrastructures. The system is engineered for on-premises AI workflows, which allows enterprises to better manage sensitive data without the risks associated with cloud computing. Extended testing with the vLLM framework has shown that this supercomputer can produce approximately 864 tokens per second output throughput, proving its capability in managing extensive and demanding open-source AI projects.

Support for Advanced AI Development

Furthermore, the ET900N G3 supports NVIDIA NemoClaw workflows, enabling enterprises to construct and deploy persistent AI assistants and autonomous agents in secure local setups. This synergy with NVIDIA’s AI software stack not only simplifies the development of advanced applications but ensures that privacy and operational flexibility remain a priority for organizations investing in AI technologies.

A Comprehensive Solution for the Future

In conclusion, the ASUS ExpertCenter Pro ET900N G3 stands as a powerful tool for AI research labs, enterprise AI deployment, content creation, simulation workloads, and tomorrow's autonomous AI applications. By merging high-performance computing with usability, ASUS not only meets but anticipates the evolving demands of AI innovators across industries.

For more information on the ExpertCenter Pro ET900N G3 or to explore deployment opportunities for enhancing your organization's AI capabilities, interested parties are encouraged to visit ASUS's official website.