Nvidia Launched DGX Spark for Local AI Computing

The new hardware allows users to run complex AI models locally to avoid cloud subscription fees.

Updated on Sept. 20, 2026 in Artificial Intelligence

Isometric editorial illustration of a black, portable server hardware unit on a plain platform, representing local AI computing systems. — Nvidia has introduced the DGX Spark, a portable hardware system that allows users to run artificial intelligence models locally, bypassing cloud-based subscription fees. AI Illustration. Upload story photo >

Live Poll

Do you trust running artificial intelligence locally on your own hardware more than in the cloud?

Nvidia has introduced the DGX Spark, a portable hardware system designed to run artificial intelligence models on private infrastructure. This device enables users to move away from cloud-based AI services and their recurring monthly costs.

Why it matters

Local AI systems store user data on private equipment, providing a more secure alternative to traditional cloud data centers. By owning the hardware, users can eliminate the monthly subscription fees required by popular cloud-based AI platforms.

The Nvidia DGX Spark features 128GB of memory and weighs 2.6 pounds. New software layers now allow for GUI-based installations, replacing the need for command line interfaces.

The players

Nvidia

Nvidia is a leading technology company that specializes in the development of graphics processing units and hardware for artificial intelligence applications.

Perplexity

Perplexity is an AI-focused technology firm that has developed a Portable Computer system specifically for the DGX Spark hardware.

Microsoft

Microsoft is a multinational technology corporation that plans to integrate GB10 chips into its Surface line of laptops.

The details

The device, which launched in October 2025, represents a shift toward personal computing hardware that handles AI tasks previously reserved for remote data centers. Manufacturers are also preparing for the fall 2026 arrival of laptops equipped with GB10 silicon, including an upcoming Microsoft Surface Laptop Ultra.

The Nvidia DGX Spark launched in October 2025.
RTX Spark laptops are expected to ship in fall 2026.

The Tech Race

The transition to local hardware signifies a departure from the reliance on centralized cloud data centers for AI processing power. This shift positions hardware manufacturers to compete directly with software providers by offering users full control over their own computing environments.

Users can gain privacy for their data by hosting AI models on their own physical hardware instead of relying on external cloud providers. This move allows owners to stop paying monthly subscription fees, though it requires a significant upfront investment for the device.

The takeaway

Transitioning to local AI hardware can provide significant data privacy benefits for users who handle sensitive information. Consumers should weigh the $4,699 cost of the device against the long-term savings of avoiding recurring monthly AI subscription fees.