As AI moves from experimentation to production, infrastructure requirements are rising fast. Larger models, multimodal pipelines, and real-time applications demand systems that can deliver high compute density, fast data movement, and consistent reliability.

Gooxi G6860T5 is built for this shift—an 8U AI server designed to handle intensive training, inference, and mixed workloads with efficiency and scale.

High-Density Compute

Gooxi G6860T5 supports up to 2 4th/5th Gen Intel® Xeon® Scalable processors, with up to 350W per CPU. Combined with 32 DDR5 slots running up to 5600MT/s, the platform provides the memory bandwidth and compute performance required for large-scale AI workloads.

PCIe 5.0 further boosts data throughput, helping eliminate bottlenecks in data-intensive scenarios and enabling smooth multi-task processing.

Optimized for GPU Acceleration

The system supports up to 8 GPUs across multiple form factors, allowing flexible configuration depending on workload and deployment needs.

A direct CPU–GPU connection reduces latency and improves data exchange efficiency. Multiple interconnect topologies can be configured to match different performance requirements, making the platform suitable for model inference, fine-tuning, training, and multimodal processing.

To maintain stability under sustained load, the server uses a multi-zone cooling design with dynamic fan control, ensuring reliable 24/7 operation.

Flexible Storage and Expansion

Gooxi G6860T5 offers hybrid storage across SATA, SAS, and NVMe. The front supports up to 12 drives, including NVMe options for high-speed access, while onboard M.2 and Slimline interfaces provide additional flexibility.

With 13 PCIe slots, the system can be easily expanded with networking, acceleration, and storage controllers—supporting a wide range of deployment architectures.

Built for Reliability

Designed for enterprise environments, the system includes:

• Hot-Swap power supplies with redundancy options
• Hardware-level security features such as TPM 2.0 and chassis intrusion detection
• Redundant and modular components for easier maintenance and higher uptime

It also supports standard remote management interfaces, enabling efficient monitoring, diagnostics, and large-scale operations.

Ready for Production AI

Gooxi G6860T5 is compatible with mainstream GPU accelerators and supports a wide range of use cases, including AI model training and inference, cloud applications, virtualization, and data analytics.

With its 8-GPU capacity, scalable architecture, and reliable design, Gooxi G6860T5 provides a solid foundation for deploying and scaling AI workloads in real-world environments.