PowerEdge model guide · GB300 NVL72 · Rack-scale AI

Dell PowerEdge XE9712: configuration and pricing factors

The PowerEdge XE9712 GB300 NVL72 configuration is a rack-scale AI platform built around interconnected Grace CPUs and Blackwell Ultra GPUs. Evaluate the rack as an integrated system, including its network, cooling, software and installation requirements.

Explore XE9712 pricing considerations

Configuration covered

NVIDIA GB300 NVL72

GPU scale

Up to 72 Blackwell Ultra GPUs

Host processors

36 NVIDIA Grace CPUs

Cooling

Direct liquid cooling

Platform overview based on Dell’s published documentation. Check the current specification and exact configuration before using it in a purchase comparison.

Workload and platform selection

Start with the model size, training or inference objective, concurrency and performance target. Confirm that the application can use the rack’s scale-up topology and memory resources. Smaller workloads should also be evaluated against alternative PowerEdge platforms; a larger GPU count is not a substitute for a workload sizing exercise.

Define the complete system scope

Define the exact XE9712 generation and NVL configuration, rack infrastructure, compute and switch trays, network interfaces and integration scope. This guide covers the GB300 NVL72 configuration; do not assume specifications from another XE9712 revision describe the same system. Request the complete component list and commissioning acceptance criteria.

Liquid cooling and facility readiness

Direct liquid cooling must be planned with the facility, rather than priced as an afterthought. Confirm who supplies the CDU, facility water connections, heat rejection and electrical distribution. Validate the selected rack’s floor loading, access route, clearances and service procedures before committing to a delivery plan.

Networking, storage and software

Distinguish the NVLink fabric inside the rack from the network connecting racks, storage and users. Document switch and adapter requirements, optics and cabling, storage performance and management access. Agree on software versions, cluster commissioning and operational handover responsibilities.

PowerEdge XE9712 pricing: what to compare

For an XE9712 price comparison, confirm that each proposal covers the same GB300 NVL72 configuration and rack integration. List network and storage equipment, software subscriptions, support, cooling and commissioning separately. Use the same currency and United States or Canada delivery and support assumptions; a GPU-only figure is not the price of a commissioned rack.

Keep these costs separate

  • Configured compute, rack, networking and storage equipment.
  • Facility power, cooling, installation and commissioning.
  • Support, software subscriptions and recurring hosting or operating costs.

PowerEdge XE9712: buyer questions

No. This guide covers the GB300 NVL72 configuration in Dell’s referenced documentation. Confirm the exact platform generation and accelerator configuration on every specification and bill of materials instead of relying on the XE9712 name alone.

The configured rack, included networking and storage, support term, software, liquid-cooling equipment, commissioning and delivery scope determine what is being compared. Keep one-time equipment and facility costs separate from recurring operating costs. This guide does not confirm a selling price or availability.

The GB300 NVL72 XE9712 configuration is a rack-scale system with up to 72 interconnected GPUs. Dell lists the XE8712 GB200 NVL4 as a four-GPU compute module for IR7000 integration. Compare complete deployment designs and workload fit rather than treating the two names as equivalent purchase units.

Manufacturer references

Continue planning