Redefining Next-Gen AI Computing Density and Thermal Limits with 6th Gen AMD EPYC™ CPUs and Advanced 2P DLC Solution

DALLAS, Sept. 22, 2026 /PRNewswire/ — AEWIN will showcase its world-first flagship servers powered by AMD EPYC™ 9006 series Server CPUs with integrated two-phase direct liquid cooling (2P DLC) solution at Taiwan Expo USA in Dallas from September 24–26, 2026. Designed for next-generation intensive AI and HPC infrastructure, the new flagship lineup includes the BG18-A10710, a 1U AI Server engineered to enable high-density heterogeneous computing power and optimized thermal efficiency from an ultra-compact chassis and the BG58-A20810, a 5U AI Server, supporting up to 16 high-performance accelerators for the most demanding AI and compute-intensive workloads.

As generative AI, large language model (LLM) training and inference, and high-performance computing (HPC) continue to drive demand for greater computing density, thermal management has become critical challenges for next-generation data centers. AEWIN addresses these challenges by integrating cutting-edge 2P DLC solutions into its latest EPYC-based server platforms. It utilizes a low-boiling-point coolant phase-change mechanism that leverages the heat generated during liquid-to-vapor transition to significantly improve heat transfer efficiency while reducing overall energy consumption.

As an industry-leading 1U AMD EPYC-based AI Server features advanced 2P DLC technology, the BG18-A10710 provides optimized thermal management to achieve ultra-high computing density and rich I/O within 1U chassis. It supports single Venice SP7 CPUs with up to 600W TDP, two PCIe Gen6 SW FHFL accelerators, 16 DDR5 RDIMM/MRDIMM memory slots with high speeds of up to 12,800 MT/s, and eight E1.S Gen6 NVMe SSD bays for high-speed storage. Combined with OCP DC-SCM2 system management via ASPEED AST2700 BMC and redundant 2,700W 80 PLUS® Titanium power supplies, the BG18-A10710 delivers a reliable and powerful foundation for high-density AI inference, HPC, and next-generation data center deployments.

For large-scale parallel computing, the BG58-A20810 is a flagship 5U AI Server powered by dual AMD EPYC™ 9006 SP8 Server CPUs with 400W TDP, supporting next-gen DDR5/MRDIMM memory subsystems for maximum throughput and bandwidth. Its advanced configuration supports up to 16 PCIe Gen6 NVIDIA RTX Pro 6000 professional GPUs integrated with 2P DLC, enabling ultimate thermal management for extreme GPU density within a single chassis. Additional configurations support up to eight air-cooled PCIe Gen6 accelerators, providing flexibility for diverse AI training, inference, HPC, and other highly compute-intensive workloads.

Powered by 6th Gen AMD EPYC™ processors and advanced 2P DLC technology, the BG18-A10710 and BG58-A20810 are designed to address the increasing computing power and thermal demands of next-generation AI infrastructure while enabling greater computing density and optimized PUE (Power Usage Effectiveness). AEWIN is ready to offer samples for technical evaluation, helping customers unleash peak computing performance while meeting ESG carbon-reduction goals to build efficient, stable, and sustainable AI infrastructure.

Exhibition Information:
Date: September 24–26, 2026
Location: Kay Bailey Hutchison Convention Center Dallas

View original content:https://www.prnewswire.com/news-releases/aewin-unveils-flagship-server-series-in-a-world-premiere-302883496.html

SOURCE AEWIN