❄️
Direct Liquid Cooling
CPUs, all 8 GPUs, and NVLink switches are directly liquid-cooled — enabling B200 at 1000W TDP in a 4U chassis while reducing data centre cooling energy costs by up to 45%.
UK’s trusted IT infrastructure partner since 2003
8× B200/H200 GPUs
NVIDIA Blackwell generation
1.44TB GPU Memory
8× B200 HBM3e NVLink pool
45% Cooling Saving
vs air-cooled alternatives
4U Form Factor
vs 6U XE9680 — denser rack
12× PCIe Gen5 Slots
For InfiniBand and DPUs
3000W PSU (B200)
Titanium efficiency rating
✅ Authorised Reseller
📦 Free UK Delivery
🔒 Full Warranty
📞 24/7 Support
Download Official Datasheet
PDF · Official manufacturer document
DellSKU: PE-XE9680L
Dell PowerEdge XE9680L Liquid-Cooled AI Server
Blackwell-Ready. Liquid-Cooled. 4U Dense.
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.
- Up to 8× NVIDIA HGX B200 (180 GB, 1000W) or H200 (141 GB, 700W) with NVLink full-mesh
- Direct liquid cooling for CPUs, GPUs, and NVLink switches — 45% lower energy cost vs air
- 4U form factor vs 6U XE9680 — 33% higher GPU density per rack with liquid cooling
- Dual Intel Xeon Scalable 5th Gen (up to 64 cores) — up to 4 TB DDR5 @ 5600 MT/s
- Up to 12× PCIe Gen5 x16 slots for dense InfiniBand and DPU connectivity
- 3000W Titanium PSUs for B200 GPU configuration; full Redfish/iDRAC9 management
Get Pricing — Speak to Our Team
Request a Quote
Competitive pricing · Response within 4 hours
Key Features
Everything you need in one device
🚀
NVIDIA Blackwell Ready
Supports NVIDIA HGX B200 (180 GB HBM3e) — Blackwell's FP8 performance exceeds H100 by 30× for inference and 4× for training, making the XE9680L a platform for next-generation AI scale.
📦
Denser Rack Footprint
4U vs XE9680's 6U means 33% more GPU nodes per rack — at 8 GPUs per node, a standard 42U rack can house nearly 70% more GPUs than with air-cooled XE9680 servers.
🔗
NVLink Full-Mesh
8× B200 GPUs share 1.44 TB of NVLink-addressable HBM3e memory at full mesh bandwidth — enables single-node training of models in the trillion-parameter class.
📡
Dense PCIe Expansion
12× PCIe Gen5 x16 slots (vs 10 on XE9680) support dual-port 400GbE/800GbE InfiniBand adapters, GPUDirect storage NICs, and BlueField-3 DPUs simultaneously.
📊
iDRAC9 Datacenter
Full iDRAC9 Datacenter with GPU telemetry, thermal monitoring for liquid cooling loop, CloudIQ predictive health, and OpenManage Enterprise fleet management.
About the PowerEdge XE9680L
The Dell PowerEdge XE9680L is the liquid-cooled evolution of Dell's flagship AI server platform — delivering NVIDIA Blackwell GPU performance in a denser, more energy-efficient 4U form factor. By bringing direct liquid cooling to CPUs, all eight GPUs, and the NVLink switch fabric, the XE9680L supports the 1000W NVIDIA B200 GPU that exceeds the thermal limits of air-cooled infrastructure.
With 8× NVIDIA B200 (180 GB HBM3e), the XE9680L delivers 1.44 TB of NVLink-addressable GPU memory — enabling training and inference of trillion-parameter scale models on a single node. NVIDIA Blackwell architecture brings a 30× inference performance uplift over H100 for FP8 workloads, and 4× faster training throughput, making the XE9680L the most capable single-node AI platform in the Dell portfolio.
The liquid cooling infrastructure reduces data centre cooling energy costs by up to 45% compared to air-cooled alternatives, while enabling the higher-TDP GPUs required for Blackwell performance. Dell's direct liquid cooling covers CPUs, all eight GPUs, and the NVLink switches — ensuring the thermal headroom needed for sustained peak GPU utilisation. The 4U chassis fits more nodes per rack than the 6U XE9680, increasing total rack-level GPU density by approximately 33%.
Technical Specifications
📄 Download Datasheet (PDF)Use Cases
Ideal for a wide range of deployments
🧬
Trillion-Parameter Models
1.44 TB NVLink HBM3e (8× B200) provides the memory required for trillion-parameter LLM training on a single node — with 30× FP8 inference uplift over H100.
🏗️
High-Density AI Clusters
4U chassis packs more nodes per rack — enabling ultra-high GPU-density AI factory deployments where space and power are constrained.
⚡
Power-Constrained Datacentres
Direct liquid cooling enables full B200 1000W TDP utilisation in data centres without air-cooling headroom — 45% lower cooling energy delivers measurable OpEx savings.
🔬
HPC & Scientific Computing
B200 FP64 performance and large HBM3e capacity support molecular dynamics, climate simulation, and genome sequencing at unprecedented scale per node.
FAQ
Frequently Asked Questions
Why Servnet
Why buy from Servnet?
Trusted UK IT reseller since 2003 — supplying businesses of all sizes with genuine, competitively priced technology.
🏆
Authorised UK Reseller
Servnet is an authorised IT reseller with direct access to leading product lines at competitive pricing.
💬
Expert Pre-Sales Advice
Our engineers will help you select the right product and advise on compatibility with your existing infrastructure.
🚚
Fast UK Delivery
Most products ship from UK stock. Express delivery available for urgent deployments.
🔒
Genuine, Warranted Kit
All products are genuine, brand-new, with full manufacturer warranty and RMA support.
📋
Configuration Services
We can pre-configure your hardware before shipping — reducing on-site deployment time and cost.
📞
24/7 Support
Post-sale support from our certified engineers. We're here if anything goes wrong.
