ASUS has introduced the XA P2N E2, a 2U server built on NVIDIA MGX architecture for agentic AI and reinforcement learning. The system pairs an NVIDIA Vera CPU with up to 2 RTX PRO 6000 Blackwell server GPUs. Unveiled at the OCP Global Summit, the machine targets high compute tasks including tool calling, code execution, and sandboxed processing.
The XA P2N E2 uses the modular NVIDIA MGX chassis design to speed up AI deployment in enterprise server racks. Inside the 2U form factor, the server combines an NVIDIA Vera processor with 2 RTX PRO 6000 Blackwell server GPUs. ASUS designed this compute configuration specifically to handle agentic artificial intelligence pipelines. These setups require fast responses for automated reasoning, background code execution, tool calling routines, and secure sandbox simulations.
ASUS also displayed several other enterprise compute platforms during the event. The lineup includes the XA VR721 E3 built on the Vera Rubin NVL72 platform, alongside the XR R1I12B M1BHFS, which pairs air cooling for the CPU with liquid cooling for the GPU block. For standard high density enterprise tasks, the company showed the RS700A E14B R16U, a 1U chassis powered by dual socket AMD EPYC 9006 processors. The display concluded with the ESC8000A E13P, an air cooled server designed for multi GPU compute setups on AMD silicon.
