• 8 Dual-Width GPU Massive Computing Power Stellar S800 tower workstation supports up to 8 dual-width GPUs, delivering extraordinary parallel computing power. It is built for heavy AI workloads including hundreds-of-billion-parameter large language model pre-training, supporting LLaMA, GPT and other mainstream LLMs.
• Massive Expandable Memory Capacity This tower AI server features up to 32 DDR5 DIMM memory slots, offering tremendous expandable RAM capacity. It efficiently handles huge datasets, complex scientific simulation and large model training tasks with reduced memory bottlenecks.
• Dual-CPU Platform & Domestic GPU Compatibility Adopts dual-CPU architecture and supports multi-GPU parallel processing with domestically produced GPUs. It realizes independent and controllable computing infrastructure, ideal for IT innovation projects and secure enterprise AI computing deployment.
• Ideal for Scientific Computing & Professional Rendering Perfect for molecular dynamics, fluid simulation, genomics research and film-grade offline rendering. It can work as a cloud rendering farm backend, meeting high-demanding workloads for universities, research institutes and media production enterprises.
• Versatile Enterprise-Grade AI Infrastructure 123L tower form factor provides ample internal space for high-end hardware installation and heat dissipation. It serves as core computing base for large model pre-training, AI fine-tuning, prototype verification, and protects data privacy via fully local offline computing.

Stellar S800 is a powerful tower AI workstation built as an eight-card computing powerhouse. Supporting up to 8 dual-width GPUs and built on dual-CPU platform, it delivers massive parallel computing performance for heavy AI and HPC workloads. It comes with up to 32 DDR5 DIMM slots for ultra-large memory expansion.

It supports multi-GPU parallel computing with domestic GPUs, enabling independent and controllable IT innovation computing infrastructure. This workstation is capable of pre-training large-scale billion-parameter LLMs such as LLaMA and GPT. It also excels at scientific computing tasks including molecular dynamics, fluid simulation and genomics, as well as film-grade offline rendering and cloud rendering farm deployment.

With a 123L tower chassis for reliable heat dissipation, it is the ideal core computing power base for research labs, AI enterprises and media studios to run large model training, scientific simulation and professional rendering locally.