Dell and AMD Expand AI Infrastructure From Single Nodes to Full Racks

Dell Technologies is expanding its Dell AI Factory with an AMD-based PowerEdge server, modular AI infrastructure configurations and the AMD Helios Rack-scale Solution from Dell. The additions are designed to support deployments ranging from a single compute node to full rack-scale clusters.
- Dell introduced the AMD EPYC-based PowerEdge M8825 server.
- The M8825 offers up to 512 cores and 1,024 threads per node.
- New Dell AI Platform with AMD configurations use AMD Instinct MI350P PCIe GPUs.
- The AMD Helios Rack-scale Solution can deliver up to 2.9 exaflops with 72 MI455X GPUs.
- Dell is offering deployment options from a single node to a full rack-scale system.
A denser server for compute-intensive workloads
At SC26, Dell introduced the Dell PowerEdge M8825, an air-cooled server based on AMD EPYC processors. According to Dell, the platform is designed to help enterprises bring agentic AI and other compute-intensive workloads into production while continuing to use environments they already operate.
Dell says the server provides up to 512 cores and 1,024 threads per node, with thermal, power delivery and operational characteristics optimized for the platform. The company describes it as a practical way for organizations to expand AMD-based AI capacity without immediately moving to a full rack-scale deployment.
Dell notes that these figures are based on preliminary product specifications from November 2026.
Modular infrastructure for starting small and scaling
Dell is also expanding the Dell AI Platform with AMD through new modular configurations built on the Dell PowerEdge XE7745 and R7725 servers. The systems use AMD Instinct MI350P PCIe GPUs and add validated support for AMD Enterprise AI Suite, Red Hat OpenShift Container Platform and Red Hat AI Enterprise.
The platform combines Dell PowerEdge servers, Dell PowerSwitch networking, Dell PowerScale storage, the Dell AI Data Platform, AI software and tailored services. Dell presents this as a full-stack, validated infrastructure solution rather than a collection of independently sourced components.
According to Dell, customers can begin with a single compute node, reuse that infrastructure as workloads move into production, and add GPUs, nodes, storage or networking in increments as demand grows. The approach is intended to reduce the need to rebuild systems from scratch or operate separate, one-off infrastructure silos.
A rack-scale option for frontier workloads
For organizations seeking higher compute capacity, Dell is offering the AMD Helios Rack-scale Solution from Dell. The company says it can deliver up to 2.9 exaflops of performance with 72 AMD Instinct MI455X GPUs.
The solution is paired with Dell PowerRack for networking and storage, along with Dell's end-to-end deployment and lifecycle services. Dell says this gives customers a single-vendor path to a validated, integrated rack-scale AI system, with the aim of reducing deployment complexity and accelerating time to value.
Dell frames the expanded portfolio as an open approach to AI infrastructure, giving customers choices in silicon, cooling, rack style and deployment scale while retaining a single partner for integration and support. The company says the options are intended for organizations starting with a smaller deployment, scaling quickly or pursuing frontier infrastructure.
Dell's performance figure for the Helios solution is based on AMD internal analysis, and the company says actual results are subject to change.
Dell Technologies: Dell Technologies and AMD Offer More Ways to Deploy AI, From First Node to Full Rack


