GPU / AI Servers

Dell PowerEdge XE

Purpose-built AI servers supporting NVIDIA H100, H200, B200, AMD MI300X, and Intel Gaudi3 — from 8-GPU training giants to 2U dense inference nodes. Enterprise-validated for production AI workloads.

← XE Overview|Get a Quote

Dell PowerEdge

PowerEdge XE9680

Dell's flagship 6U, 8-way GPU server supporting NVIDIA H100/H200 SXM5, AMD MI300X, or Intel Gaudi3 — purpose-built for LLM training, inference, and HPC at enterprise scale.

PowerEdge XE9680L

Dell's next-generation 4U liquid-cooled AI server supporting NVIDIA B200 (180 GB) or H200 (141 GB) GPUs with NVLink — delivering peak Blackwell performance in a dense, efficient chassis.

PowerEdge XE9640

Dell's densest 2U GPU server — 4× NVIDIA H100/H200 SXM5 GPUs with direct liquid cooling. Delivers up to 2× GPU core density per rack versus standard air-cooled configurations.

XE Server Comparison

Spec	XE9680 (Flagship)	XE9680L (Blackwell)	XE9640 (Dense)
Form Factor	6U Air-Cooled	4U Liquid-Cooled	2U Liquid-Cooled
GPU Count	8×	8×	4×
GPU Options	H200/H100/MI300X/Gaudi3	B200/H200	H100/H200 SXM5
Max GPU Memory	1.12 TB (8× H200)	1.44 TB (8× B200)	564 GB (4× H200)
GPU Interconnect	NVLink / Infinity Fabric	NVLink full-mesh	NVLink full-mesh
System Memory	Up to 4 TB DDR5	Up to 4 TB DDR5	Up to 4 TB DDR5
PCIe Gen5 Slots	10×	12×	4×
Cooling	Air (to 35°C)	Direct Liquid	Direct Liquid
Best For	Training + diverse GPU	Blackwell / max perf	Dense inference

Need Help Choosing the Right XE Server?

Our team can help size GPU memory, NIC bandwidth, and storage capacity for your specific AI workload — LLM training, fine-tuning, inference, or HPC.

Talk to an Expert