- Demand is growing for infrastructure capable of supporting increasingly autonomous and complex workloads
- Efficient AI infrastructure depends on more than compute power alone, it requires optimization across power, cooling, connectivity, and automation
- AI factories are emerging as integrated environments for deploying, managing, scaling, and governing AI, with ASUS showcasing solutions from enterprise AI platforms to rack-scale AI POD deployments
AI is entering a phase where infrastructure matters as much as the models themselves. At ASUS AI Tech 2026 in Seoul, ASUS and its ecosystem partners explored how the shift from generative AI to agentic and physical AI is transforming what organizations need from their AI infrastructure. From performance and efficiency to scalability and governance, the event highlighted the technologies and partnerships shaping the future of AI infrastructure.
AI is entering a new phase
During his keynote, CH Hsieh, ASUS Corporate Vice President of R&D Center, emphasized that AI infrastructure must evolve alongside AI itself. As AI workloads become increasingly diverse and complex, organizations need infrastructure that can support them efficiently and at scale. He also highlighted a key question facing the industry: not how to build bigger and faster data centers, but how much it actually costs to produce a token. This shift is driving greater focus on infrastructure optimization across power, cooling, connectivity, and automation.
ASUS AI Tech 2026
Every signal. Every watt. Every degree. Every second
As AI workloads grow more complex, infrastructure efficiency is becoming as important as raw computing performance. Throughout the event, ASUS emphasized that power, cooling, connectivity, and automation should not be viewed as separate technologies, but as interconnected elements of a modern AI infrastructure strategy. By reducing inefficiencies across the stack, organizations can improve performance, optimize resources, and maximize the value of their AI investments.
In his keynote, CH Hsieh summarized this approach with the principle: “Every Signal. Every Watt. Every Degree. Every Second.” Together, these factors help determine how efficiently AI systems operate and how effectively organizations can scale future AI workloads.
From infrastructure to AI factories
As AI becomes increasingly embedded in business operations, organizations need environments that can support the entire AI lifecycle, from deployment and management to ongoing optimization and governance. This shift is driving the evolution from traditional infrastructure toward AI factories: integrated environments that combine computing, storage, networking, deployment software, AI operations, and governance into a unified framework.
At ASUS AI Tech 2026, examples included enterprise AI server platforms, HGX training systems, and rack-scale ASUS AI POD deployments powered by the latest NVIDIA Vera Rubin NVL72 server platform, designed to support large-scale AI training and supercomputing workloads.
AI factories also emerged as a key theme for the next stage of AI adoption. Rather than treating AI as a series of isolated projects, organizations are increasingly looking to build scalable environments that can support deployment, operations, governance, and continuous optimization across the entire AI lifecycle. This approach provides a foundation for long-term AI growth and enterprise-wide AI adoption.
From enterprise AI to rack-scale infrastructure
Supporting AI at scale requires more than a single infrastructure approach. Different workloads place different demands on computing resources, from enterprise inference and model training to large-scale AI environments. Throughout the event, ASUS showcased a portfolio designed to meet these evolving requirements, helping organizations build infrastructure that supports AI today while preparing for future growth.
Highlights included solutions spanning enterprise AI servers, advanced training platforms, and rack-scale ASUS AI POD deployments. Together, these offerings illustrate ASUS's strategy of delivering the right infrastructure for every stage of the AI journey, from individual workloads to large-scale AI environments.
As AI continues to evolve, so too will the infrastructure required to support it. From the rise of agentic and physical AI to the growing demand for scalable AI factories, organizations are facing new challenges and opportunities on their AI journey.
Throughout ASUS AI Tech 2026, one message remained clear: AI infrastructure is no longer defined by compute power alone. Organizations must also address power efficiency, cooling, connectivity, and automation to reduce infrastructure losses and maximize AI performance.
By combining enterprise AI servers, HGX training platforms, and rack-scale AI POD deployments powered by NVIDIA Vera Rubin NVL72, ASUS is building infrastructure that supports AI workloads ranging from enterprise inference to large-scale training and supercomputing.
ASUS (2357.TW) is an innovative technology leader delivering the world’s most comprehensive AI solutions across Infrastructure, Physical AI, AI PC/Devices, and Engineering the AI Advantage. Guided by its “Ubiquitous AI. Incredible Possibilities” strategy, ASUS is bringing enterprise-to-edge AI to life. The ASUS portfolio also includes the ROG and ProArt sub-brands, which serve gamers and creators worldwide.
https://asus.com
