Squadron NVL72

Complete liquid-cooled rack, 18 compute trays · Rack-scale AI

NVIDIA GB300 NVL72 rack-scale AI. Complete liquid-cooled rack, 18 compute trays with 36× NVIDIA Grace and liquid cooled, requires a cdu design, configured with Evolve III around your workload.

Evolve III Squadron NVL72 server

Squadron NVL72 server

Why organisations choose it

A complete NVIDIA GB300 NVL72 rack with 18 compute trays.

Liquid cooled and requires a coolant distribution unit (CDU). Planned through a consultation.

Available models

1 models in this family.

  • E3-DLB2-CB3

    Form factor
    Complete liquid-cooled rack, 18 compute trays
    Processor
    36× NVIDIA Grace
    Accelerator
    72× NVIDIA Blackwell Ultra, 20 TB HBM3E (system total)
    Storage
    Designed with your data platform

Technical specification

Platform

Form factor
Complete liquid-cooled rack, 18 compute trays
Processor
36× NVIDIA Grace
GPU / accelerator
72× NVIDIA Blackwell Ultra, 20 TB HBM3E (system total)

Storage & cooling

Storage
Designed with your data platform
Cooling
Liquid cooled, requires a CDU

Specification notes

Memory, storage capacity, networking and power are configured to the workload and confirmed on quotation.

Discuss this configuration

Tell us what you want to build.

Share your device requirements, target quantities and market. Our team will help you explore suitable products and the next steps for your project.

  • Distributors and resellers taking a range to their own market.
  • Education, government and institutional buyers.
  • Enterprise buyers and organisations with product requirements.