Featured Image Credit: Carl Court / Staff / Getty
On Monday (13 October), Nvidia announced that it has begun shipping the DGX Spark, which is reportedly the world's smallest AI supercomputer.
If you wanted to get your hands on one of these, you're looking at an eye-watering price tag of $3,999.
In a symbolic gesture, Nvidia CEO Jensen Huang personally hand-delivered the first units to none other than Elon Musk.
“The exchange was a connection to the supercomputer’s origins, as Musk was among the team that received the first NVIDIA DGX-1 supercomputer from Huang in 2016,” the company explained.
Advert
The original DGX-1 went to a small startup called OpenAI, where it helped create ChatGPT.
According to the tech giant, DGX Spark is primarily designed to handle a wide range of AI workloads in a compact form factor that is 'powerful enough to accelerate agentic and physical AI development.'
It is equipped with the whole of NVIDIA’s AI stack including GPUs, CPUs, networking, CUDA libraries and more.
"By combining breakthrough performance with the reach of the NVIDIA ecosystem, DGX Spark transforms the desktop into an AI development platform," NVIDIA stated.
“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, kickstarting the AI revolution,” said Jensen Huang, founder and CEO of NVIDIA. “DGX-1 launched the era of AI supercomputers and unlocked the scaling laws that drive modern AI. With DGX Spark, we return to that mission — placing an AI computer in the hands of every developer to ignite the next wave of breakthroughs.”
Despite its compact size, DGX Spark delivers genuinely impressive performance. The company says it provides a petaflop of AI computing power - that's one quadrillion operations per second - equipped with 128 GB of unified memory.
Powered by NVIDIA's GB10 Blackwell Superchip, the system can run inferences on AI models containing up to 200 billion parameters and fine-tune models with up to 70 billion parameters.
“In addition, DGX Spark lets developers create AI agents and run advanced software stacks locally,” said NVIDIA.
For context, that means it can handle some of the largest and most complex AI models currently in development, all from a desktop machine.
Beyond Musk, tech giants including Microsoft, Google, Hugging Face, and Meta have received DGX Spark units for testing and optimising their AI tools and platforms.
“DGX Spark allows us to access peta-scale computing on our desktop,” said Kyunghyun Cho, professor of computer and data science at the NYU Global Frontier Lab. “This new way to conduct AI research and development enables us to rapidly prototype and experiment with advanced AI algorithms and models - even for privacy- and security-sensitive applications, such as healthcare.”
