New Delhi: Nvidia has recently announced big news and shared the launch of its DGX Spark, the world’s smallest AI supercomputer. The tech firm reckons AI workloads are quickly outgrowing the memory and software capabilities of the PCs, workstations, and laptops that millions of developers rely on, forcing teams to shift work to the cloud or local data centres. As the latest class of the complex, DGX Spark delivers AI performance and 128GB of unified memory in a compact desktop form factor, giving the developer the power to run inference on the AI models up to 200 billion parameters and fine-tune models of up to 70 billion parameters locally.
DGX Sparks brings together the Nvidia AI platform, including the GPUs, CPUs, and networking, CUDA libraries, and the AI software stack into a system small enough for the lab or an office, yet powerful enough for agentic and physical AI development. DGX Spark system delivers up to 1 petaflop of AI performance, accelerated by an Nvidia GB10 Grace Blackwell Superchip, Nvidia ConnectX-7 200Gb/s networking, and Nvidia NVLink-C2C technology, providing 5x the bandwidth of the fifth-generation PCIe with 128GB of CPU-GPU coherent memory.
Other early recipients of DGX Spark, including Anaconda, Cadence, ComfyUI, Docker, Hugging Face, JetBrains, LM Studio, Meta, Microsoft, Ollama and Roboflow, are testing, validating, and optimising their tools, software, and models for DGX Spark. Research organisations around the world, including the NYU Global Frontier Lab, previewed DGX Spark.
