The XA P2N-E2 integrates up to two NVIDIA RTX PRO 6000 Blackwell GPUs, prioritizing high single-thread performance and memory efficiency to keep AI agents responsive during demanding evaluation cycles. This release anchors a broader ASUS infrastructure push, which includes the XA VR721-E3 AI POD—a liquid-cooled rack system packing 36 NVIDIA Vera CPUs and 72 Rubin GPUs to support large-scale training and inference environments.
Beyond specialized hardware, ASUS is diversifying its portfolio to accommodate varied data center requirements. The company showcased Intel Xeon 6-based platforms for modular deployment and new AMD EPYC 9006-series systems designed for high-density, data-intensive tasks. By combining these compute options with NVIDIA’s Spectrum-X networking and CMX context memory, ASUS aims to provide a standardized, full-stack architecture capable of scaling from single-server setups to massive AI factories.
Comments (0)
No comments yet. Be the first!