Nvidia sells tiny new computer that puts big AI on your desktop

Nvidia sells tiny new computer that puts big AI on your desktop

Nvidia has officially unveiled the DGX Spark, a groundbreaking desktop AI computer that is set to transform how developers approach artificial intelligence tasks. Priced at $4,000, this compact powerhouse boasts an impressive one petaflop of computing performance and 128GB of unified memory, all contained within a desktop-friendly design. The DGX Spark aims to cater to the growing needs of AI developers by providing substantial integrated memory, enabling the execution of larger AI models that traditional consumer GPUs cannot handle. Orders for the device will commence on Wednesday, October 15, via Nvidia's website, with additional availability through select retail partners in the United States. Formerly known as "Project DIGITS," the DGX Spark represents Nvidia's strategic move to establish a new category of desktop workstations specifically tailored for AI development. This innovation addresses a critical challenge faced by many AI developers: the limitations of standard PCs and workstations often drive them to rely on cloud services or data centers for their work. Despite its promising capabilities, the market for desktop AI workstations remains uncertain, especially when considering the upfront investment compared to cloud solutions that offer flexible pay-as-you-go models. However, the DGX Spark is designed to support larger AI models locally, accommodating up to 200 billion parameters and fine-tuning models with as many as 70 billion parameters without the need for remote infrastructure. The applications for this technology are vast, including the ability to run advanced open-weight language models and media synthesis tools, such as AI-driven image generators. Users can customize models like Black Forest Labs' Flux.1 for image generation and leverage Nvidia’s Cosmos Reason language model for vision search and summarization tasks. Furthermore, the system is optimized for building chatbots using the Qwen3 model. Weighing only 2.65 pounds and measuring 5.91 x 5.91 x 1.99 inches, the DGX Spark is an engineering marvel that operates on Nvidia's GB10 Grace Blackwell Superchip. It features cutting-edge ConnectX-7 200 Gb/s networking and NVLink-C2C technology, which delivers five times the bandwidth of PCIe Gen 5, along with its robust 128GB of unified memory that enhances both system and GPU performance.

Sources : Ars Technica

Published On : Oct 17, 2025, 04:19

Startups
India's Ambitious Electronics Initiative: A Game-Changer for Local Manufacturing

In a bold move to enhance the global footprint of India's electronics industry, Union IT Minister Ashwini Vaishnaw unvei...

Business Today | Mar 30, 2026, 09:25
India's Ambitious Electronics Initiative: A Game-Changer for Local Manufacturing
Streaming
YouTube's Chief Confirms Creators' Loyalty Amid Streaming Competition

Neal Mohan, the CEO of YouTube, recently expressed confidence that the platform's most successful creators will remain l...

TechCrunch | Mar 29, 2026, 21:50
YouTube's Chief Confirms Creators' Loyalty Amid Streaming Competition
AI
Revolutionizing Stadium Safety: How AI is Changing Crowd Management for IPL 2026

As the Indian Premier League (IPL) gears up for 2026, ensuring the safety of enthusiastic fans has become a pressing pri...

Business Today | Mar 30, 2026, 07:45
Revolutionizing Stadium Safety: How AI is Changing Crowd Management for IPL 2026
AI
Sarvam AI Launches 'Chanakya' to Address High-Security AI Demands

Sarvam, a burgeoning AI startup from India, is set to enhance its portfolio with a new initiative named 'Chanakya.' This...

Business Today | Mar 30, 2026, 05:20
Sarvam AI Launches 'Chanakya' to Address High-Security AI Demands
Automotive
The Rise of Robotaxis: Navigating Challenges and Funding Surges in Autonomous Transport

Welcome to the forefront of transportation innovation, where robotaxis are rapidly gaining traction. Waymo has recently ...

TechCrunch | Mar 29, 2026, 16:30
The Rise of Robotaxis: Navigating Challenges and Funding Surges in Autonomous Transport
View All News