Unveiling Compal's Powerful AI Servers: Meet The SX420-2A, SX220-1N, And SX224-2A Revolutionizing HPC At GTC 2025

In the rapidly evolving landscape of artificial intelligence and high-performance computing (HPC), staying ahead of the curve is essential for enterprises looking to leverage cutting-edge technology. At the GTC 2025, Compal Electronics unveiled an impressive lineup of AI servers—namely the SX420-2A, SX220-1N, and SX224-2A—each designed to meet the rigorous demands of modern computing environments. Built on NVIDIA’s MGX architecture, these servers promise enhanced performance, adaptability, and efficiency for a wide range of AI and HPC applications. In this article, we delve into the features, specifications, and potential impact of these servers on the data center landscape.

Table of Contents

Takeaways

Compal’s new server lineup includes the SX420-2A, SX220-1N, and SX224-2A, all leveraging NVIDIA’s MGX architecture.
The SX420-2A supports up to 8 NVIDIA RTX PRO 6000 GPUs, making it ideal for demanding AI-HPC applications.
The SX220-1N features NVIDIA’s GH200 Grace Hopper Superchip, enhancing performance through faster memory speeds and massive bandwidth.

Overview of Compal’s New Server Platforms

## Overview of Compal’s New Server Platforms
Compal Electronics, a leader in technology innovation, has recently taken a giant leap forward in the realm of enterprise-level AI and high-performance computing (HPC) with the unveiling of its latest server platforms at GTC
2025. The company’s trio of advanced servers—the SX420-2A, SX220-1N, and SX224-2A—are all meticulously designed on NVIDIA’s MGX architecture, tailored specifically to meet the demanding requirements of AI and HPC applications.

The SX420-2A, an impressive 4U AI Server, stands out with its capability to support up to eight RTX PRO 6000 Blackwell GPUs, enhancing its performance for diverse AI-HPC applications. Each GPU is equipped with 96GB of cutting-edge GDDR7 memory and employs a passive cooling design, ensuring reliable operation even during the most intensive workloads, making it ideal for data centers looking to optimize their resources.

On the other hand, the SX224-2A, a versatile 2U AI Server, incorporates the NVIDIA MGX with an AMD x86 platform, allowing for flexible configurations that can be adjusted to support both AI workloads and graphics-heavy tasks. This unique design caters to businesses that require customizable performance to effectively address their computational demands.

Lastly, the SX220-1N, also a 2U server, is powered by NVIDIA’s GH200 Grace Hopper Superchip alongside NVLink-C2C technology. This remarkable combination facilitates incredibly fast memory speeds and provides vast bandwidth capabilities, which are vital for managing large-scale applications efficiently.

Since its founding in 1984, Compal has earned a reputation for its forward-thinking approach to technology. With these new server offerings, the company is not only reinforcing its commitment to innovation but is also positioning itself as a key player in the AI and HPC sectors, significantly enhancing the capabilities of data centers across various industries.

Technical Specifications and Performance Insights

When examining the specifications of Compal’s new server platforms, it’s clear that each model is engineered with unique features aimed at maximizing performance for enterprise needs. The SX420-2A is pivotal for AI-driven data centers, where its architecture supports a robust GPU arrangement, thereby ensuring optimal throughput for intensive AI workloads. Its advanced cooling system is crucial for maintaining performance stability under pressure, while the support for up to 96GB per GPU enables handling of larger datasets efficiently. In contrast, the SX224-2A promotes adaptability in its design, effectively meeting the dual demands of AI and graphics applications, which is increasingly vital as businesses seek equipment that can serve multiple purposes without sacrificing efficiency. Lastly, the SX220-1N leverages cutting-edge technology with its Grace Hopper Superchip, bringing forth not only higher memory speeds but also enhanced processing capabilities for heavy computational tasks. This diverse range of server solutions underscores Compal’s commitment to providing top-tier hardware that meets the evolving demands of today’s digital landscape.