ASUS has announced the XA GB721-E2, built on the NVIDIA GB300 NVL72 rack-scale system, marking a major step forward in AI and data center infrastructure. Designed for large-scale model training, high-throughput inference, and HPC workloads, the solution delivers trillion-parameter performance with 36 NVIDIA Grace CPUs and 72 NVIDIA Blackwell Ultra GPUs in a single NVLink domain for ultra-fast, low-latency processing.
The system integrates seamlessly within a 48U NVIDIA MGX-compliant rack, featuring liquid cooling, modular architecture, and balanced storage to ensure high efficiency, scalability, and sustainability. It also supports rapid deployment enabling full-rack installation in just 30 minutes through the ASUS Infrastructure Deployment Center (AIDC).
Complemented by ASUS Control Center and end-to-end Professional Services, ASUS provides customers with complete lifecycle support from design and integration to long-term maintenance. This combination of automation, energy-efficient cooling, and adaptable infrastructure enables organizations to accelerate AI innovation while maintaining operational reliability.
With this launch, ASUS reinforces its leadership in next-generation computing and sustainable technology, empowering enterprises and research institutions to build data centers optimized for performance, precision, and sustainability.
Leave a comment