AEWIN Unveils AMD EPYC 9006 SP7 Servers for AI and Cloud Workloads

Four server platforms target different workloads

AEWIN’s portfolio includes the BG18-A10710, BG28-A10710, BX28-A10710 and BS28-A10710. The company describes the systems as purpose-built platforms for AI workloads, enterprise applications, scalable storage and cloud infrastructure.

For AI computing, the BG18-A10710 is a 1U AI server that supports up to two single-width 600W GPUs. It also integrates two-phase direct liquid cooling, or 2P DLC, for CPU and GPU thermal management.

The BG28-A10710 is a 2U AI server that accommodates up to four full-height, full-length GPUs. According to AEWIN, PCIe 6.0 connectivity supports demanding AI training, inference and accelerated computing workloads.

General-purpose and storage systems add platform options

The BX28-A10710 is a 2U general-purpose server designed to combine compute and storage capabilities. Its listed configuration includes support for two PCIe 6.0 slots, one OCP 3.0 slot and up to 12 U.2 NVMe SSDs.

For performance-critical storage workloads, the BS28-A10710 is a 2U storage server supporting up to 16 E3.S Gen6 SSDs. AEWIN says the system is designed to provide high-density storage capacity and high-bandwidth connectivity for data-intensive applications.

These configurations give the portfolio different roles within infrastructure deployments. The AI systems focus on GPU-accelerated work, while the general-purpose and storage systems address enterprise, cloud and data-intensive requirements.

AEWIN prepares systems for upcoming shipments

AEWIN, a member of the Qisda Group and a server and network appliance provider, said the BG28-A10710 and BX28-A10710 highlight its platform readiness as it prepares to bring its latest EPYC servers to market in alignment with AMD’s EPYC 9006 SP7 processors.

The systems were demonstrated at AMD in Depth: EPYC "Venice" in Las Vegas, following their preview on November 5. The announcement does not provide specific shipment dates or pricing.

AEWIN says its approach translates the capabilities of AMD EPYC 9006 SP7 Server CPUs into platforms with high-performance computing power, high-speed connectivity and flexible configurations. The company also points to the wider technology capabilities of Qisda Group ecosystem partners as part of its effort to support rack-scale AI infrastructure.

For customers, the announcement signals a choice of server designs for AI, general-purpose computing and storage rather than a single platform. AEWIN says the portfolio is intended to balance performance with cost-effective scalability as organizations build infrastructure for changing workload requirements.