GIGABYTE

Designed around 8 × NVIDIA HGX H200 GPU module (SXM) — giving 900 GB/s GPU‑to‑GPU bandwidth via NVLink + NVSwitch, suitable for large‑scale distributed GPU compute.

HPC/AI Server - AMD EPYC™ 9004 - 5U DP NVIDIA HGX™ H100 8-GPU 4-Root Port Powering the Next Generation of Server Architecture and Energy Efficiency

5U HPC/AI server with AMD EPYC 9005/9004 CPUs, NVIDIA HGX™ H200 8‑GPU, designed for AI, HPC, AI training, and AI inference workloads.

5U dual-CPU HPC/AI server with 5th/4th Gen Intel® Xeon® Scalable processors, 8× NVIDIA HGX™ H100 GPUs, optimized for AI, HPC, AI training, and AI inference.

HPC/AI Server - 5th/4th Gen Intel® Xeon® Scalable - 4U DP 8 x PCIe Gen5 GPUs

High-density 4U GPU server for AI, visual computing, HPC, AI training, and AI inference with dual EPYC CPUs and up to 10 GPUs.

High-performance 3U AI server with dual processors and up to 4 GPUs, optimized for AI training, inference, and HPC workloads.

HPC/AI Server - 5th/4th Gen Intel® Xeon Scalable - 2U DP 8 x PCIe Gen5 GPUs (with PCIe switches)

AI & HPC-ready 2U Arm server powered by Ampere® Altra Max, supporting up to 4× PCIe Gen4 GPUs — ideal for AI training, inference, and high-performance computing workloads.

NTS Elite APEX 2U HPC & AI Compute System – dual-socket AMD EPYC™ 9005/9004 server with up to 4 PCIe Gen5 GPUs, optimized for AI, HPC, visual computing, and GPU-intensive workloads.

High-performance 2U server for AI, visual computing, HPC, AI training, and AI inference workloads.