Skip to content
HomeSolutionsGPU ServersDell PowerEdge XE9680 — 8x AMD Instinct MI300X (192GB HBM3)
Nexus Compute

Dell PowerEdge XE9680 — 8x AMD Instinct MI300X (192GB HBM3)

Flagship air-cooled MI300X platform with 1.5TB of coherent HBM3 per node.

Full manufacturer warrantyAuthorized channel48-hour quote

We help you choose, configure, and deliver the right system — no obligation.

Dell PowerEdge XE9680 — 8x AMD Instinct MI300X (192GB HBM3) — Nexus Compute enterprise hardware
Dell PowerEdge XE9680 — 8x AMD Instinct MI300X (192GB HBM3) hardware detail 1
Dell PowerEdge XE9680 — 8x AMD Instinct MI300X (192GB HBM3) hardware detail 2
Dell PowerEdge XE9680 — 8x AMD Instinct MI300X (192GB HBM3) hardware detail 3

Configuration at a Glance

Accelerator8x AMD Instinct MI300X OAM, 192GB HBM3 each (1.5TB total)
GPU InterconnectAMD Infinity Fabric, all-to-all (6.4TB/s bisection)
CPUDual Intel Xeon Scalable (4th/5th Gen)
System MemoryUp to 4TB DDR5 (32 DIMM slots)

Tailored per engagement. Full technical overview below.

Configuration Options

Core specifications for this system. Every component is configurable to your workload — request a quote for a tailored build.

GPU / Accelerator

8x AMD Instinct MI300X OAM, 192GB HBM3 each (1.5TB total)

Processor

Dual Intel Xeon Scalable (4th/5th Gen)

Memory

Up to 4TB DDR5 (32 DIMM slots)

Storage

Up to 16x E3.S NVMe or 8x 2.5in NVMe/SAS/SATA

Overview

The Dell PowerEdge XE9680 pairs eight AMD Instinct MI300X OAM accelerators — 1.5TB of total HBM3 — with dual Intel Xeon Scalable CPUs in a serviceable air-cooled 8U chassis. Nexus Compute specifies, configures, and burn-in tests each system on ROCm, then delivers it through authorized Dell channels with full enterprise warranty.

Who This Solution Is For

Enterprises standardizing on Dell for AMD-accelerated AI
Teams running memory-heavy LLM inference and fine-tuning
Organizations that prefer air-cooled deployment without facility plumbing
Buyers wanting a Tier-1 OEM platform with global support

Business Benefits

Massive memory per node

1.5TB of HBM3 holds very large models on a single server, reducing sharding and operational complexity.

Air-cooled serviceability

Gold-grade fan cooling deploys into standard high-density racks without direct-liquid plumbing.

Tier-1 OEM assurance

Sourced through authorized Dell channels with enterprise warranty and global field support.

Typical Business Use Cases

1

Large-context LLM inference and serving

2

Foundation-model fine-tuning on ROCm

3

Consolidating large models onto fewer nodes

4

Memory-bandwidth-bound HPC and AI workloads

Industry Applications

AI & Machine LearningFinancial ServicesHigher Education & ResearchGovernment & DefenseHPC

Technical Overview

Built on the Dell PowerEdge XE9680 8U platform with eight 750W MI300X OAM modules linked by 4th-gen AMD Infinity Fabric for all-to-all GPU communication. Dual Intel Xeon Scalable processors, up to 4TB of DDR5, and front-serviceable E3.S NVMe feed the accelerators, with 2800W Titanium redundant power.

Accelerator8x AMD Instinct MI300X OAM, 192GB HBM3 each (1.5TB total)
GPU InterconnectAMD Infinity Fabric, all-to-all (6.4TB/s bisection)
CPUDual Intel Xeon Scalable (4th/5th Gen)
System MemoryUp to 4TB DDR5 (32 DIMM slots)
StorageUp to 16x E3.S NVMe or 8x 2.5in NVMe/SAS/SATA
NetworkingUp to 8x PCIe Gen5 NICs for GPUDirect RDMA
Form Factor8U rackmount, air-cooled
Power2800W Titanium redundant, hot-swap
WarrantyDell ProSupport, enterprise-grade (configurable term)

Specifications are indicative and configured to each engagement. Request a quote for a configuration tailored to your requirements.

Warranty, Support & Fulfillment

Every system ships from an authorized channel, configured and tested, with the documentation enterprise buyers need — backed by warranty and a dedicated account team.

Enterprise Warranty

Full manufacturer warranty with optional on-site, next-business-day support and extended coverage.

Authorized Channel

Sourced through Tier-1 distribution and OEM partners — never grey market. Asset & warranty records included.

Lead Time & Deployment

48-hour quotes, then configured, burn-in tested, and delivered on a committed schedule.

Nationwide Fulfillment

Coordinated logistics, rack-and-stack, and delivery wherever your infrastructure lives.

Frequently Asked Questions

How large a model fits on one XE9680?

With 1.5TB of pooled HBM3, a single node serves models well beyond 100B parameters in common precisions without sharding across machines. We size precision and batch settings to your target latency during scoping.

Is MI300X production-ready on ROCm?

Yes. ROCm supports mainstream frameworks and inference engines such as vLLM and PyTorch. We validate your stack on the configured system before delivery.

Can it run in a standard server room?

The XE9680 is air-cooled and deploys into high-density racks with adequate power and airflow. We confirm exact power and thermal requirements and can advise on colocation.

Hardware Assistance

Configure the Dell PowerEdge XE9680 — 8x AMD Instinct MI300X (192GB HBM3) with Nexus Compute

Tell us your requirements and a hardware specialist will help you specify, configure, and quote the right system — typically within two business days. No obligation.