Description
Custom RTX graphics cards for AI servers
These custom high-memory RTX configurations are intended for GPU compute workloads where memory capacity and card dimensions matter. The listed options combine expanded VRAM with a dual-slot blower cooler for integration into a compatible server or workstation.
Available configurations
| RTX 3090 | 48GB VRAM; custom dual-slot blower configuration. |
|---|---|
| RTX 4090 | 48GB VRAM; custom dual-slot blower configuration. |
Confirm your system requirements
Tell us which model and quantity you need, your chassis, power supply and intended software workload. We confirm the exact card configuration, physical clearance, power connections, airflow requirements and software compatibility in the quotation. A dual-slot form factor alone does not establish compatibility with every chassis or multi-GPU layout.
These are custom configurations. The listed memory capacities and cooling arrangements should not be read as standard NVIDIA retail specifications. Availability, final specifications, testing and the applicable commercial warranty are confirmed for the quoted cards.
Request a quotation
Choose your model and quantity and send an enquiry for the GPUs alone or as part of a complete NodeX server. The quotation states the confirmed configuration, price, lead time and warranty. Submitting an enquiry does not place an order.
See our Warranty Policy and Terms of Use. Mandatory consumer rights remain unaffected.

