ASUS Launches Advanced XA P2N-E2 Server with NVIDIA Vera CPU for AI Applications

ASUS Introduces XA P2N-E2 with NVIDIA Vera CPU at OCP 2026

ASUS has unveiled its latest innovation, the XA P2N-E2 server, during the OCP Global Summit 2026 in San Jose, California. This cutting-edge technology is designed specifically to meet the needs of agentic AI and reinforcement learning workloads. Built on the NVIDIA MGX architecture and powered by NVIDIA Vera CPUs, the XA P2N-E2 aims to streamline the execution layer of AI applications by efficiently managing a variety of complex tasks.

The XA P2N-E2 boasts a compact 2U design, capable of supporting up to two dual-slot NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. This configuration allows users to combine high-performance CPU capabilities with versatile GPU acceleration, making it suitable for demanding applications in AI environments. With high single-thread CPU performance and energy-efficient, high-bandwidth memory, the server is optimized to keep AI agents responsive, even under strenuous execution and evaluation tasks.

At the OCP Global Summit, alongside the XA P2N-E2, ASUS presented an extensive range of products aimed at enhancing AI infrastructure. This includes the XA VR721-E3 AI POD, which integrates a staggering 36 NVIDIA Vera CPUs and 72 NVIDIA Rubin GPUs, all maintained within a fully liquid-cooled architecture for efficient large-scale AI training and inference.

The advancements do not stop there. ASUS has launched the XR R1I12B-M1BHFS, a server that utilizes NVIDIA HGX Rubin NVL8 architecture with a combination of direct-to-chip liquid cooling and air cooling for Intel® Xeon® 6 processors. This diverse offering underscores ASUS's commitment to addressing the varying compute density, cooling architectures, and deployment requirements essential for today's complex AI workloads.

Broader Implications of AI Infrastructure

ASUS is redefining what it means to construct an AI factory. Their strategy goes beyond simply increasing compute power; it encompasses high-speed networking, scalable context memory, and enhanced data infrastructure to support the intricacies of AI environments. By leveraging NVIDIA's technologies, ASUS is bringing together a comprehensive ecosystem that allows for increased flexibility and scalability.

For instance, the integration of NVIDIA's DSX reference design with the NVIDIA Omniverse DSX Blueprint aids in the creation of digital twins, allowing businesses to validate architecture and refine AI factory designs prior to implementation. This approach not only accelerates the design process but ensures that operational efficiencies are realized from the outset.

Moreover, the introduction of essential technologies at ASUS’s booth, such as NVIDIA HGX Rubin NVL8 for accelerated computing and the NVIDIA Spectrum-X based Ethernet switching helps to facilitate high-speed networking while maintaining responsive data movements essential for AI performance.

With an increasing demand for AI applications, NVIDIA Groq 3 LPX stands out, providing ultra-fast token generation while supporting the needs of agentic systems. This addition complements NVIDIA Vera Rubin NVL72, creating a comprehensive architecture that ensures seamless performance across compute, networking, context memory, inference, and storage elements.

Choosing the Right Architecture

ASUS also highlighted diversity in server architecture at the event, showcasing platforms powered by both Intel and AMD. The Intel showcase included models like the RS700-E12-RS4U and RS720-E12-RS12U, which are designed to accommodate the diverse needs of cloud computing, AI, and high-performance computing. Providing modular deployment options and tool-less serviceability, these servers are adaptable to various applications, including agentic AI and data-heavy workloads.

The AMD display featured the ASUS RS700A-E14B-R16U, which integrates 6th Gen AMD EPYC 9006-series processors, addressing enterprise-level resilience and high-density requirements. The unique design incorporates advanced thermal management, ensuring stability and peak performance even in challenging environments.

ASUS's commitment to delivering excellent thermal management alongside high-density capabilities allows organizations to deploy AI systems without needing specialized cooling setups, further enhancing operational flexibility.

Across the OCP Global Summit 2026, ASUS is proving that it is at the forefront of AI infrastructure technology. It combines system engineering, accelerated computing, and networking to provide adaptable solutions for AI, cloud, and high-performance computing workloads. Attendees are invited to explore the full range of the ASUS AI infrastructure portfolio at Booth #B9.

For global clients, ASUS ensures that their servers are available worldwide, though specific product configurations may differ by region, prompting inquiries to local representatives for more details.