Back to stories
Industry

Dell PowerRack Ships GB200 NVL72 as a Turnkey Rack, Live in 6.5 Hours

Michael Ouroumis2 min read
Dell PowerRack Ships GB200 NVL72 as a Turnkey Rack, Live in 6.5 Hours

Dell launched PowerRack at Dell Technologies World 2026 on Monday — a turnkey, rack-scale platform that bundles PowerEdge XE9712 servers paired with NVIDIA GB200 NVL72 compute, eight PowerSwitch SN6600-LD network switches, PowerFlex storage, and a 220kW liquid-cooling distribution unit as a single validated SKU. Dell says a unit can go from delivery dock to live workloads in six and a half hours.

Inside the rack

Each PowerRack collapses compute, networking, storage, and cooling into one pre-cabled system. Compute is built on liquid-cooled PowerEdge XE9712 nodes wired into NVIDIA's GB200 NVL72 rack-scale platform. The networking layer is eight PowerSwitch SN6600-LD switches — Dell-branded versions of NVIDIA's Spectrum SN6600-LD — delivering more than 800Tb/sec of east-west switching capacity. Storage runs on PowerFlex, supporting block, file, and object on the same fabric. Cooling is handled by the new PowerCool CDU C7000, which fits 220kW of liquid-cooling capacity into a 4U chassis with relaxed inlet-temperature requirements.

Storage scales for the model pipeline

Dell expanded ObjectScale, its S3-compatible object store, with 245TB drives that triple per-rack density. The platform now indexes billions of unstructured files for AI dataset creation and runs GPU-accelerated analytics that Dell claims yield up to 6x faster query performance. ObjectScale also added on-premises support for foundation models from Google, OpenAI, Palantir, and SpaceXAI — a signal that enterprise buyers want frontier weights running next to their data rather than over the wire.

AI Factory hits 5,000 customers

Dell's NVIDIA-co-developed AI Factory added 1,000 customers last quarter to reach 5,000 total, with Eli Lilly, Honeywell, and Samsung among the disclosed new logos. The company also debuted Deskside Agentic AI for local agent execution that keeps sensitive data off external clouds, and deepened a Mistral collaboration aimed at sovereign deployments. A new AI-driven partner platform — promising deal registration in minutes rather than days and dynamic real-time pricing — is slated for August 2026.

What changes for buyers

The pitch is sharp. Instead of integrating XE9712 servers, Spectrum switches, PowerFlex pools, and CDUs themselves, enterprises receive one validated, pre-cabled rack designed for a same-day stand-up. That compresses procurement, integration, and acceptance testing into a single contract — and tightens Dell's grip on the on-prem GB200 NVL72 buildout against HPE's Cray-branded racks and Supermicro's NVL72 SKUs. For teams trying to move workloads off rented clouds without owning a hardware lab, the rack-as-a-SKU model lowers the bar to deploy frontier compute behind the firewall, with cooling and switching engineered as part of the BOM rather than a post-purchase scramble. The networking and storage delivery dates in late 2026 are the gating items — until then, PowerRack lands as a compute-led product, with the full-stack pitch arriving in the second half.

Learn AI for Free — FreeAcademy.ai

Take "AI for Business: Practical Implementation" — a free course with certificate to master the skills behind this story.

More in Industry

Google Ships Antigravity 2.0: A Standalone Agent Platform That Retires the Gemini CLI
Industry

Google Ships Antigravity 2.0: A Standalone Agent Platform That Retires the Gemini CLI

At I/O 2026, Google relaunched Antigravity as a standalone agent-orchestration platform with a CLI that replaces the Gemini CLI, an SDK, Managed Agents in the Gemini API, and an enterprise tier — all defaulting to Gemini 3.5 Flash.

10 min ago2 min read
Alibaba's Zhenwu M890 Claims 3x Nvidia's H20, Ships With Qwen 3.7-Max
Industry

Alibaba's Zhenwu M890 Claims 3x Nvidia's H20, Ships With Qwen 3.7-Max

Alibaba's chip unit T-Head unveiled the Zhenwu M890 AI accelerator, claiming 3x the agentic-inference performance of Nvidia's H20 and 144GB of HBM3, paired with the Qwen 3.7-Max model — a full domestic stack as US export controls tighten.

1 hours ago2 min read
Cohere Buys Reliant AI to Build 'North for Pharma' Sovereign Biopharma Agents
Industry

Cohere Buys Reliant AI to Build 'North for Pharma' Sovereign Biopharma Agents

Cohere has acquired Montréal- and Berlin-based biopharma startup Reliant AI, folding its automated research workbench into the North platform to launch 'North for Pharma' — its second acquisition in a month as it doubles down on sovereign, regulated-industry AI.

3 hours ago2 min read