On December 10, Sugon unveiled a detailed account of its investor relations engagements, addressing market inquiries concerning orders for its super-node 640 product, its compatibility with the H200, and its collaborative efforts with NVIDIA. Sugon highlighted that its scaleX640 model, equipped with cutting-edge hardware architecture, cooling, and power supply technologies, stands at the forefront of global integration and claims the top spot in China for super-node comprehensive performance. As the scale of large model parameters continues to grow and high-throughput inference clusters are increasingly deployed, the advantages of this product are set to become even more pronounced, boasting vast market potential. Furthermore, the Sugon scaleX640 super-node embraces an open AI computing architecture, ensuring compatibility with a wide range of mainstream domestic and international AI accelerator cards available on the market. This allows users to tailor their accelerator chip selection based on specific needs, thereby achieving an AI computing solution that embodies both 'software-hardware synergy and ecosystem compatibility'.