Need AI? Dell backs up the truck and tips out servers, storage, blueprints
Vendor leans on Nvidia tie-up so hard you can hear the GPUs squeak
SC25 Dell continues to push itself as a one-stop shop for enterprise AI infrastructure with a wave of products and services, including updates to servers, storage, and software to expand its offerings.
At the SC25 High Performance Computing event in St. Louis, Missouri, Dell was keen to talk up its partnership with Nvidia as a central part of its pitch to enterprise customers struggling to get AI up and running.
The pair already combined forces on the Dell AI Factory with Nvidia, a fully integrated AI system, but this is now supported by Dell's Automation Platform, which offers centralized delivery and management of IT operations.
The aim of this system is to help businesses rapidly deploy AI agents, the new must-have capability in the AI world, according to Varun Chhabra, Dell's Senior Vice President of Infrastructure.
"Customers will be able to access a curated catalog of validated workload blueprints through the Dell automation platform, and IT teams will have access to this library of AI blueprints that can then be deployed on AI Factory with Nvidia configurations," he said.
Another development Dell was keen to trumpet is integration between Nvidia's Dynamo inference framework and the PowerScale and ObjectScale storage platforms via the NIXL communication library.
This enables KV cache offloading, which moves the large key-value (KV) cache data for processing large language models from GPU memory to cheaper storage, reducing GPU memory usage and improving performance.
According to Dell, its own testing of this configuration delivered a one-second time to first token (TTFT), even with a full context window of 131,000 tokens, compared with the standard vLLM configuration, which took over 17 seconds.
On the hardware side, Dell is adding new systems such as the PowerEdge XE8712 server, available from next month, designed to be delivered as part of a self-monitoring rack-scale platform, purpose-built for AI and HPC.
A standard rack of these can be configured with up to 144 of Nvidia's GB200 GPUs, the firm said, along with an updated Integrated Dell Remote Access Controller (iDRAC), an Integrated Rack Controller (IRC) for advanced thermal control, and a new rack-mounted coolant distribution unit (CDU).
That rack controller provides visibility and automation across every layer of infrastructure, and with the CDU, that includes the liquid cooling for the servers, according to Dell.
"In liquid-cooled environments, the Integrated Rack Controller continuously monitors for leaks, responding within seconds to isolate risk and safeguard uptime," said Deania Davidson, Senior Director of Product Management for Dell's AI Server and Networking. "It can even detect a leak as small as 20 microliters, giving operators the confidence that even the tiniest leaks won't go unnoticed."
Dell also said that its PowerScale platform will soon be available as software that customers can run on qualified Dell servers like the PowerEdge R7725xd, while PowerScale also aims to deliver greater performance with support for parallel NFS, which enables parallel IO operations to a cluster of PowerScale storage systems.
The Dell PowerEdge XE8712 server will be available globally in December, while the AI workload blueprints are in tech preview, and the software-defined PowerScale and PowerScale parallel NFS support will be globally available in 2026. ®