Dell ships first liquid-cooled storage and networking for AI racks
Dell Technologies has started shipping a 2U, 32-drive PowerEdge server and matching networking rack—both cooled entirely by liquid. The move gives AI data centers a single cooling blueprint and lets operators repurpose storage as workloads change. It also opens the door to tapping stranded power sources for distributed AI.
Editor, Lazyfounder

30 SEC SUMMARY
- Dell released a 2U, 32-drive PowerEdge server with 100% liquid cooling—the first built for AI racks.
- The system runs Dell Exascale Storage, letting operators switch between file, block, object, or parallel file storage on the same hardware.
- Dell also shipped a pretested, liquid-cooled networking rack that cuts deployment time from weeks to hours.
- Compute, storage, and networking now share the same cooling and power specs, so data centers can be built as single liquid-cooled units.
- Dell’s software control plane can manage up to 1,000 small sites with spare power, turning them into distributed AI capacity.
TABLE OF CONTENTS
A 100% liquid-cooled storage node for AI racks
Dell Technologies has rolled out a 2U PowerEdge server that packs 32 drives and 100% liquid cooling—the first storage system built for AI racks that ditches air cooling entirely. The server runs Dell Exascale Storage, a software layer that lets operators switch between file, block, object, or parallel file storage without touching the hardware.
Storage software that adapts to workload shifts
AI workloads are changing fast. Training clusters need high-throughput file storage; inference jobs often call for low-latency block storage; agentic tasks can flip between the two in the same hour. Dell’s answer is to separate the storage software from the hardware. Operators can deploy whatever persona they need today and reassign the same drives tomorrow, avoiding vendor lock-in.
George Papadopoulos, Dell’s Senior Vice President of Product and Engineering, says the split speeds up bringing new PowerEdge generations to market.
Networking rack arrives pretested and liquid-cooled
Dell also shipped a pretested, liquid-cooled networking rack that includes switches, optics, and cabling. The rack arrives fully configured, cutting deployment time from weeks to hours. Ihab Tarazi, Dell’s Senior Vice President and Chief Technology Officer for AI, Compute and Networking, says the move reflects how AI clusters now scale from tens of thousands to hundreds of thousands of GPUs on a single fabric.
Because compute, storage, and networking all use the same cooling and power specs, data-center operators can now design a single liquid-cooling loop for everything, rather than separate systems for each component.
Stranded power becomes AI fuel
AI clusters are so power-hungry that new data centers are being planned with 100% liquid cooling. Tarazi says customers are already asking for it, because air cooling can’t handle the heat from thousands of GPUs running flat-out.
That same power density is creating opportunities beyond traditional data centers. Retail stores, industrial sites, and even methane-capture projects have spare power that isn’t being used. Dell’s software control plane is designed to manage up to 1,000 of these stranded-power sites, turning them into distributed AI capacity.
A shift from components to complete systems
Earlier this year, Dell added semantic layers and knowledge graphs to its AI Data Platform to sharpen data context for enterprise models. The liquid-cooled storage and networking gear released today tackles the physical side—heat and power—that comes with running those models at scale.
In May, Dell partnered with AMD to build rack-scale AI infrastructure for enterprises, a deal that fits with its current shift toward selling complete, integrated systems rather than individual components.
What this means
Lazyfounder analysis — our interpretation, not reported fact.
Dell is betting that AI factories are too power-hungry for air cooling and too unpredictable for fixed storage appliances. By offering liquid-cooled storage and networking that share the same specs as compute, Dell gives operators a single blueprint for new data centers. At the same time, the software-defined storage layer lets them repurpose hardware as AI workloads change—without buying new gear.
Key takeaways
- The 2U PowerEdge node is the first 100% liquid-cooled storage system designed for AI racks.
- Dell Exascale Storage lets operators flip between file, block, object, or parallel file storage on the same server.
- A pretested liquid-cooled networking rack now ships fully configured, reducing deployment time from weeks to a few hours.
- Compute, storage, and networking now use identical cooling and power specs, simplifying data-center design.
Related on Lazyfounder
Sources
- SiliconANGLE · 2026-11-17
Dell bets liquid-cooled, software-defined storage can keep AI factories fed - SiliconANGLE · 2026-11-18
From store outlets to methane leaks, every scrap of power is now fair game for AI compute
This story is an original summary drafted with AI by Lazyfounder from the reporting listed above and checked by automated validation. Facts are attributed to their original publishers; sections marked as analysis are Lazyfounder's. Where a source is in another language, facts were machine-translated and quotations are reported, not reproduced. Read the original coverage via the links, and see our AI policy and corrections policy.
About the author
Editor, Lazyfounder
Tarun Mottlia edits LazyFounders, covering Indian startups, funding rounds, AI and product launches. Every story on the site is AI-assisted and checked against its cited sources before publication.
More stories by Tarun MottliaGet the LazyFounder Brief
Startup, funding and AI news in a five-minute read. Join the early-access list.


