NVIDIA 940-54242-0000-000 DGX Spark Founders Edition USA

NVIDIA 940-54242-0000-000 DGX Spark Founders Edition

Overview

The NVIDIA 940-54242-0000-000 is the DGX Spark Founders Edition (USA), a compact AI compute platform built around a 20-core Arm processor (10x Cortex-X925 cores) paired with a 256-bit memory interface. The 4TB NVMe M.2 drive ships with hardware self-encryption enabled — relevant for deployments handling sensitive inference workloads or regulated data. The system runs NVIDIA DGX OS out of the box, eliminating driver-stack setup time for teams moving directly into AI development or edge inference workflows. Display output is via HDMI 2.1a; wireless connectivity includes Bluetooth 5.3 with LE support. The single RJ-45 Ethernet connector handles network integration. This is a USA-market unit.

Compatibility

Designed to run NVIDIA DGX OS and software stacks built for the DGX Spark platform. The HDMI 2.1a output is compatible with any standard HDMI 2.1 display. Bluetooth 5.3 with LE is compatible with standard BT 5.x peripherals. The 4TB NVMe M.2 self-encrypting drive operates within the DGX Spark's native storage controller — replacement or expansion drives must be validated against the DGX Spark storage specification. Ethernet connectivity via the single RJ-45 port is compatible with standard 1GbE/10GbE switches depending on the onboard NIC spec; consult NVIDIA's DGX Spark network requirements before provisioning switch ports. HDMI multichannel audio output routes through the HDMI 2.1a connector — no separate analog audio output is available on this unit.

Installation Notes

Unit is USA-market only (940-54242-0000-000); confirm region before ordering — DGX Spark SKUs vary by geography and power configuration. The 4TB NVMe with self-encryption requires key management provisioning before first use in regulated environments. NVIDIA DGX OS is pre-installed; validate your target software stack against the DGX OS version shipped before deployment. No discrete GPU add-in card slot is implied by available evidence — workload planning should account for the integrated compute architecture of the DGX Spark platform.