Infrastructure

Edge Servers

Short-depth and micro data centre platforms for near-premise compute

Edge servers sit between devices and the cloud: telecom cabinets, retail back offices, hospital equipment rooms and factory server closets. They typically emphasise short chassis depth, remote management and constrained power and cooling.

Solutions in this category

Representative platforms with the details engineering and procurement teams shortlist on.

Short-depth rack server installed in a small edge equipment cabinet
Dell Technologies

Dell PowerEdge XR7620

Short-depth two-socket 2U server purpose-built for edge sites running video analytics and inference.

Form factor
Short-depth 2U, two-socket
Positioning
Purpose-built for edge deployment
Target workloads
Industrial automation, video analytics, point-of-sale analytics, AI inferencing
Deployment
Edge rack or shallow cabinet
Store and branch inference
Telecom edge
Local model hosting
Box-format edge server with multiple expansion slots
Supermicro

Supermicro SYS-E403-13E-FRN2T IoT SuperServer

2.5U edge box server with a single Xeon Scalable socket, up to 2 TB DDR5 and three full-height PCIe 5.0 x16 slots.

Processor
Single-socket 4th or 5th Gen Intel Xeon Scalable
Memory
Up to 8 DIMMs, 2 TB 3DS ECC DDR5
Expansion
3× PCIe 5.0 x16 full-height full-length slots
Deployment
Edge cabinet, communications room or plant server closet
Multi-access edge computing
Industrial and retail inference
Local model serving
Edge server with hot-swap drive bays in a compact rack
Lenovo

Lenovo ThinkEdge SE455 V3

EPYC 8004 edge server supporting up to 64 cores, 768 GB of TruDDR5 memory and up to six single-wide accelerators.

Processor
AMD EPYC 8004, up to 64 cores, SP6 socket, TDP to 200 W (cTDP 225 W)
Memory
16 GB to 768 GB TruDDR5 4800/5600 MHz RDIMM, 6 DIMM slots
Accelerators
Up to 2 double-wide or up to 6 single-wide GPUs
Deployment
Edge rack, back office or plant server room
Retail and branch AI
Manufacturing inference
Distributed model serving
Rack-mounted edge AI server with multiple graphics cards installed
Advantech

Advantech AIR-520 edge AI server

4U AMD EPYC 7003 edge AI system taking up to four single-slot GPUs with a 1200 W ATX 3.0 power supply.

Processor
AMD EPYC 7003 series, options up to 64 cores
Accelerators
Up to 4 single-slot or 2 dual-slot GPUs; NVIDIA-Certified with RTX 6000 Ada
Memory
Up to 768 GB DDR4-3200
Deployment
Plant or facility rack
Multi-stream video analytics
On-premise model serving
Industrial AI aggregation

Specifications are summarised from publicly published manufacturer material and are provided for orientation only. Always confirm current figures directly with the manufacturer before purchasing or designing in.

Companies working in this category

Advantech

Hardware manufacturer

Industrial computing manufacturer supplying modular fanless edge systems such as the MIC-770 series and GPU-dense edge AI servers such as the AIR-520 for factory and infrastructure deployments.

Taipei, Taiwan

OnLogic

Hardware manufacturer

Builds configurable industrial computers, including fanless systems with Intel Core Ultra processors and expansion for Hailo accelerators or NVIDIA graphics cards.

South Burlington, Vermont, United States

Dell Technologies

Cloud & infrastructure provider

Supplies the PowerEdge XR range of short-depth, edge-focused servers for industrial automation, video analytics and inference at retail, telecom and factory sites.

Round Rock, Texas, United States

Lenovo

Hardware manufacturer

Manufactures ThinkEdge edge servers and ThinkStation workstations, including multi-GPU platforms for local model development and EPYC-based edge systems for distributed sites.

Beijing, China / Morrisville, United States

Supermicro

Cloud & infrastructure provider

Builds IoT SuperServer and edge box systems with Xeon Scalable processors and full-height PCIe expansion for accelerators, aimed at telecom, industrial and retail edge sites.

San Jose, California, United States

SUSE (Rancher)

Software provider

Originator of K3s, the lightweight CNCF-conformant Kubernetes distribution now a CNCF project, widely used to orchestrate containerised AI workloads on edge and Arm hardware.

Nuremberg, Germany

Edge Servers: common questions

When is an edge server better than the cloud?
When latency budgets are tight, when bandwidth to move raw video or sensor data is expensive, or when data residency and privacy rules require processing to stay on site.