NVIDIA GTC 2025: Compal Unveils GPU Server Platform Built on NVIDIA MGX Architecture
Including SX420-2A 4U AI, SX224-2A 2U AI, and SX220-1N 2U AI models
This is a Press Release edited by StorageNewsletter.com on March 24, 2025 at 2:01 pmAt NVIDIA GTC 2025, Compal Electronics unveiled 3 new server platforms: SX420-2A, SX220-1N, and SX224-2A.
All are built on NVIDIA MGX architecture and designed to inject powerful performance into enterprise-level AI, HPC, and high-load computing applications.
Server SX420-2A and SX224-2A, powered by NVIDIA RTX PRO 6000 Blackwell Server Edition
-
The SX420-2A 4U AI server, making its debut, is engineered on the NVIDIA MGX architecture and optimized for deep AI-HPC applications. It features a flexible rack design that supports both EIA 19 ”and ORV3 21” configurations, and it can be configured with up to 8 RTX PRO 6000 Blackwell GPUs – significantly boosting data center compute performance and resource utilization.
-
The SX224-2A 2U AI server, integrates NVIDIA MGX architecture with an AMD x86 platform, offering highly flexible configuration options. It is engineered for future compatibility and optimized for diverse computing workloads—whether AI-HPC or AI-Graphics—allowing for tailored performance adjustments.
RTX PRO 6000 Blackwell GPU delivers top-tier performance: equipped with 96GB of ultra-fast GDDR7 memory and featuring a passively cooled thermal design, it ensures stable operation under extreme loads, providing acceleration for both agentic and physical AI, as well as for scientific computing, graphics, and video applications. Combining this GPU performance, the SX420-2A and SX224-2A are undoubtedly one of the most competitive solutions in the industry.
SX220-1N: Scalable AI/HPC and Future-Proof Flexible Computing Platforms
-
SX220-1N 2U AI server is designed for giant-scale AI and HPC applications, featuring the NVIDIA GH200 Grace Hopper Superchip and employing NVIDIA NVLink-C2C technology to deliver a coherent memory pool. This enables faster memory speeds and massive bandwidth to tackle large-scale computational tasks.
This product lineup not only satisfies the rigorous requirements of diverse data center applications but also heralds a new era in enterprise computing technology.