All products
Nvidia Unveils the DGX Spark: A Compact Powerhouse for AI Computing
Nvidia has introduced the DGX Spark, a compact, ARM-based system designed to bring high-performance AI computing to the desktop. The DGX Spark is powered by Nvidia’s DGX OS, an Ubuntu Linux-based operating system optimized for GPU processing, and comes preinstalled with Nvidia’s AI software stack, including CUDA libraries and NIM microservices.
The DGX Spark is priced at $3,999, which may seem steep, but considering the cost of high-end GPUs with ample video RAM, such as the RTX Pro 6000 (around $9,000) or AI server GPUs (starting at $25,000 for a base-level H100), it represents a more affordable option. Although it may not match the performance of these high-end GPUs, the DGX Spark offers a more accessible entry point for developers and researchers working with AI models.
Performance and Capabilities
The DGX Spark’s GPU computing performance is roughly equivalent to an RTX 5070, according to The Register. However, the 5070 is limited to 12GB of video memory, which restricts the size of AI models that can be run on the system. In contrast, the DGX Spark boasts 128GB of unified memory, enabling it to run larger models, albeit at a slower speed than more powerful GPUs like the RTX 5090. For instance, running the 120 billion-parameter version of OpenAI’s gpt-oss language model requires around 80GB of memory, which exceeds the capacity of most consumer GPUs.
A Nod to the Past
Nvidia founder and CEO Jensen Huang marked the launch of the DGX Spark by personally delivering one of the first units to Elon Musk at SpaceX’s Starbase facility in Texas, echoing a similar delivery he made to Musk at OpenAI in 2016. Huang noted that the DGX-1, launched in 2016, was designed to give AI researchers their own supercomputer, and the DGX Spark represents a return to that mission. “In 2016, we built DGX-1 to give AI researchers their own supercomputer. I hand-delivered the first system to Elon at a small startup called OpenAI, and from it came ChatGPT,” Huang said in a statement.
For more information on the Nvidia DGX Spark and its capabilities, visit Here
Image Credit: arstechnica.com