• Add to Compare

NVIDIA DGX™ Spark Founders Edition

  • Powered by the GB10 Superchip
  • 1,000 TOPS of AI performance using FP4
  • NVIDIA Blackwell GPU with fifth-generation Tensor Core technology
  • NVIDIA Grace CPU implementation with 20-core high-performance Arm architecture
  • 128 GB of coherent, unified memory
  • Up to 4 TB of NVMe storage
  • Support for up to 200B-parameter large language models
  • Connectivity with NVIDIA Connect-X networking to link two DGX Spark units for up to 405B-parameter models
  • Compact design, takes up minimal desktop space
  • Plugs into standard wall outlet

NVIDIA DGX™ Spark belongs to a new class of computers designed from the ground up to build and run AI. Powered by the NVIDIA GB10 Grace Blackwell Superchip, DGX Spark delivers up to 1, 000 AI TOPS1 of performance to power AI workloads. With 128 GB of coherent unified system memory, developers can experiment, fine-tune, or inference the latest generation of reasoning AI models from DeepSeek, Meta, Google, and others of up to 200B parameters. Plus, NVIDIA ConnectX™ networking can connect two DGX Spark AI computers to enable inference on models up to 405B parameters.

To give developers a familiar experience, DGX Spark mirrors the same software architecture that powers industrial-strength AI factories. Using the NVIDIA DGX OS with Ubuntu Linux and preconfigured with the latest NVIDIA AI software stack, along with developer program access to NVIDIA NIM and NVIDIA Blueprints, developers can hit the ground running using common tools such as Pytorch, Jupyter, and Ollama to prototype, fine-tune, and inference on DGX Spark and seamlessly deploy to DGX Cloud or any accelerated data center or cloud infrastructure.

Features

NVIDIA GPU, CPU, Networking, and AI Software Technologies

NVIDIA GB10 Superchip

Unlock up to 1000 AI TOPS of FP4 precision performance powered by the NVIDIA Grace Blackwell architecture.

128 GB of Coherent Unified System Memory

Develop and test AI models with up to 200 billion parameters right at your desktop, powered by large, unified system memory.

NVIDIA ConnectX Networking Cores

High-speed NVIDIA ConnectX networking lets you link two NVIDIA DGX Spark systems to handle AI models with up to 405 billion parameters.

NVIDIA AI Software Stack

Leverage a complete full-stack platform for generative AI—featuring tools, frameworks, libraries, and pretrained models.

Workloads

Accelerate All AI Workloads

NVIDIA DGX Spark packs AI supercomputing power into a compact form factor—perfect for developers, researchers, and data scientists tackling demanding AI workloads.

Prototyping

Develop, test, and validate AI models and applications.

Powered by the NVIDIA AI software stack, NVIDIA DGX Spark offers a robust platform for developers to build, test, and validate AI models and AI-enhanced applications. When ready for final tuning or deployment, seamlessly scale to NVIDIA DGX Cloud or other NVIDIA-accelerated data center and cloud environments.

Fine-Tuning

Fine-tune AI models up to 70 billion parameters.

Enhance the performance of pretrained models by fine-tuning them on NVIDIA DGX Spark. With 128GB of unified system memory, you can fine-tune models with up to 70 billion parameters—enabling customized AI solutions tailored to your specific use cases.

Inference

Test, validate, and inference with AI models up to 200 billion parameters.

Fifth-generation Tensor Cores with support for FP4 deliver up to 1,000 TOPS of AI computing performance, combined with 128GB of system memory, accelerate inference of state-of-the-art AI models to test, validate and deploy from your NVIDIA DGX Spark.

Data Science

Accelerate end-to-end data science workflows with NVIDIA RAPIDS™.

From data preparation to training, inference, and deployment, NVIDIA RAPIDS on NVIDIA DGX Spark streamlines development with zero-code-change acceleration and familiar APIs—making it easy to speed up existing workloads, build tailored solutions, and seamlessly scale to the data center or cloud.

Edge Applications

Develop edge applications with NVIDIA AI frameworks, including Isaac™, Metropolis, and many others.

NVIDIA DGX Spark is an ideal platform for building solutions in robotics, smart cities, and computer vision. With support for NVIDIA frameworks like Isaac, Metropolis, and Holoscan, developers can harness the full power of DGX Spark to rapidly create and deploy edge applications.

Architecture NVIDIA Grace Blackwell
GPU Blackwell Architecture
CPU 20 core Arm,10 Cortex-X925 + 10 Cortex-A725
CUDA Cores Blackwell Generation
Tensor Cores 5th Generation
RT Cores 4th Generation
Tensor Performance1 1,000 AI TOPS
System Memory 128GB LPDDR5x, unified system memory
Memory Interface 256-bit
Memory Bandwidth 273 GB/s
OS DGX OS
Storage 1 or 4 TB NVME.M2 with self-encryption
USB 4x USB Type C
Ethernet 10 GbE
1x RJ-45 connector
NIC ConnectX-7 Smart NIC
Wi-Fi Wi-Fi 7
Bluetooth BT 5.3
Audio output HDMI multichannel audio output
Power Consumption 170W
Display Connectors 1x HDMI 2.1a
NVENC | NVDEC 1x | 1x
System Dimensions 150 mm L x 150 mm W x 50.5 mm H
System Weight 1.2 kg

Preliminary specifications, subject to change
1 Theoretical FP4 TFLOPS using the sparsity feature.

  • Revised web page of product spec and information won't be noticed, product colorbox printing shows the actual information of the product.
  • Above product spec is for reference only, actual spec rely on the real product and Leadtek keeps the right to alter. Each sales region will impacts the product difference, please contact your supplier for making sure the actual product information.
  • The adapter, cable and software listed on the web page are for reference only and Leadtek keeps the right to alter, revised information won't be noticed.
  • Above brand name and product name are trademark of each corresponding company.