PowerEdge XE9680 (8x AMD Instinct MI300X 192GB OAM)
Largest per-GPU memory for ROCm-based AI at massive scale
We help you choose, configure, and deliver the right system — no obligation.
Configuration at a Glance
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.
8x AMD Instinct MI300X 192GB 750W OAM
2x 5th Gen Intel Xeon Scalable, up to 64 cores each
1.5TB HBM3 coherent pool
Up to 8x U.2 NVMe or 16x E3.S NVMe
Overview
This PowerEdge XE9680 variant is configured with eight AMD Instinct MI300X 192GB OAM accelerators linked by AMD Infinity Fabric, delivering the highest GPU memory capacity in the XE line for open ROCm software stacks. Nexus Compute specifies, integrates, and tests every system, supplied with Dell ProSupport through authorized channels.
Who This Solution Is For
Business Benefits
1.5TB GPU memory pool
Eight 192GB MI300X accelerators provide 1.5TB of coherent HBM3, fitting the largest models entirely in GPU memory.
Open software stack
ROCm and PyTorch support give teams an open, portable path without proprietary lock-in.
Validated integration
Each node ships thermally and topology-validated with Infinity Fabric verified and Dell ProSupport in place.
Typical Business Use Cases
Memory-resident large model training
High-capacity generative inference
Open-stack AI research
HPC and AI converged workloads
Industry Applications
Technical Overview
The XE9680 OAM configuration interconnects eight AMD Instinct MI300X accelerators over AMD Infinity Fabric, aggregating 1.5TB of HBM3 with high bisection bandwidth for data-parallel and model-parallel training. Dual 5th Gen Intel Xeon Scalable CPUs, DDR5 memory, and PCIe Gen5 host I/O drive the platform under the open ROCm software stack with iDRAC9 management.
| GPU/Accelerator | 8x AMD Instinct MI300X 192GB 750W OAM |
| GPU Interconnect | AMD Infinity Fabric, fully connected 8-GPU |
| Total GPU Memory | 1.5TB HBM3 coherent pool |
| CPU | 2x 5th Gen Intel Xeon Scalable, up to 64 cores each |
| Memory | 32x DDR5 DIMM, up to 4TB |
| Storage | Up to 8x U.2 NVMe or 16x E3.S NVMe |
| Networking/Fabric | NVIDIA ConnectX-7 400Gb NDR InfiniBand |
| Form Factor | 6U rack, air-cooled |
| Warranty | 3-year Dell ProSupport |
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
Why choose MI300X over an NVIDIA configuration?
The MI300X offers 192GB per GPU, the largest in the XE line, plus an open ROCm software stack, making it compelling for very large memory-resident models and teams that want accelerator vendor choice.
Is my existing PyTorch code compatible?
Most PyTorch and mainstream frameworks run on ROCm with minimal change; we can advise on stack readiness and validate your key workloads during integration.
What support backs this platform?
Systems are delivered with Dell ProSupport arranged through authorized distribution, and Nexus Compute handles integration, testing, and firmware baselining beforehand.
Hardware Assistance
Configure the PowerEdge XE9680 (8x AMD Instinct MI300X 192GB OAM) 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.