As the global AI computing race reaches new heights, high-performance, scalable, and reliable AI servers have become essential infrastructure for building intelligent capabilities. Capturing this momentum, Gooxi has introduced two powerful 4U AI servers based on the Intel® Eagle Stream platform — SY4104G-G4 and AS4110G-G4. Featuring flexible configurations and exceptional scalability, these systems deliver precise computing power tailored to diverse AI workloads.

Exceptional Performance, Engineered for Intelligence

Both models support 4th and 5th Gen Intel® Xeon® Scalable processors, with each CPU offering up to 350W TDP for outstanding multi-core performance. They excel in network control, data preprocessing, and lightweight AI inference tasks.

Each server includes 16 DDR5 memory slots running up to 5600MHz, boosting memory bandwidth by 75% over the previous generation. This enables smoother, faster data throughput for AI training and inference — ensuring models are “fed” efficiently and operate without bottlenecks, forming a robust foundation for next-generation intelligent computing.

Flexible Expansion for Every AI Scenario

Recognizing that no two AI workloads are the same, Gooxi designed the two 4U systems with distinct expansion strategies:

• AS4110G-G4 maximizes compute density with an innovative CPU–GPU switch expansion design, supporting up to 10 full-height, dual-width GPUs within a single 4U chassis. This “supercomputing node in a box” is ideal for large model training, scientific research, and other heavy-duty workloads, enabling enterprises to achieve massive computing performance in compact deployments.

• SY4104G-G4 delivers all-round flexibility, supporting up to 4 dual-width GPU accelerators and 13 PCIe 5.0 slots for expansion. It also supports 10Gb network adapters and various storage modules. With 12 front 3.5-inch drives and rear NVMe U.2 modules, it enables efficient tiered data storage — perfect for AIGC content generation, virtualization, and database analytics.

Built for Stability and Continuous Uptime

In large-scale AI clusters, reliability is critical. Both servers are equipped with 8038 hot-swap fans and CRPS redundant power supplies — 1+1 redundancy for the SY4104G-G4 and 3+1 for the AS4110G-G4 — ensuring continuous, stable operation under extreme loads.

An onboard BMC chip supporting IPMI 2.0 enables remote monitoring, fault diagnosis, and maintenance without physical access, dramatically improving management efficiency in distributed environments. This guarantees uninterrupted operation for mission-critical AI services around the clock.

Powering the Future of Intelligent Computing

Gooxi’s SY4104G-G4 and AS4110G-G4 have already been deployed across leading-edge applications such as digital twins, large model training, and AIGC content creation. From single-node inference to large-scale GPU clusters, Gooxi is committed to delivering full-stack AI infrastructure solutions that go beyond hardware — empowering enterprises to lead in the era of intelligent computing and accelerating the development of smart computing ecosystem.