Page 7 - Edgecore Product Catalogue 2026

Post-Training AI Cluster for Small-Scale Deployment
AI Cluster Solution for Small-Scale Deployment

AI Server Backend Fabric GPU Cluster AS7326-56X

AIS800-64O 25GE
400GE

AGS8600
AI Server #1 AI Server #2 AI Server #3 (up to 16 units) AI Server #16
High-Performance GPU Computing for AI Workloads Frontend Fabric

Edgecore AI Server delivers high-performance GPU computing for AI and ML workloads, providing 25GE
exceptional performance and scalability for LLM training, AI inference, computer vision, and next- 200GE
generation AI infrastructure.
AS9726-32DB

400GE
1GE 1GE 1GE
OOB MGMT

Storage Fabric Storage Server Farm

Large-scale Pre-Training AI Data Center for Scalable Deployment
Large-Scale AI Data Center Solution for Scalable Deployment

AGS8600 Backend Fabric GPU Cluster N

SU (Scalable Unit) xN Spine- 1 Spine- 2
8 × AMD Instinct™ MI325X GPUs AIS800-64O
Dual AMD EPYC™ 9005 Processors 800GE
7 × Infinity Fabric™ GPU Interconnect
PCIe Gen5 & 400G Ethernet Ready
Leaf-1 Leaf-2 Leaf-3 Leaf-4 Leaf-5 Leaf-6 Leaf-7 Leaf-8
ROCm™ AI Software Ecosystem Rail Optimized
1GE
AIS800-64O
800GE break to
2X400GE AI Server #1 AI Server #2 AI Server #3 AI Server #4 AGS8600
AI Server #N

400GE
1GE
200GE
Leaf-1 Leaf-2 AS9726-32DB Leaf-4 Leaf-N Leaf-1 Spine-1
1GE
400GE
1GE 400GE
400GE
Spine-1 Spine-2 400GE Leaf-2 Spine-2

AS9736-64D AS9726-32DB AS9736-64D
Storage Fabric Storage Server Farm Frontend Fabric
6 7