A Grace Blackwell AI supercomputer on your desk.
NVIDIA DGX™ Spark belongs to a new class of computers designed to build and run AI, delivering up to 1,000 AI TOPS performance from a compact and power-efficient desktop form factor.
NVIDIA GPU, CPU and AI Software Technologies
NVIDIA GB10 Superchip
Advanced Grace Blackwell architecture delivers exceptional AI compute performance in a compact desktop form factor.
128GB Unified Memory
Run and fine-tune large AI models locally with ultra-fast coherent system memory architecture.
ConnectX Networking
Connect multiple DGX Spark systems together for larger AI models and distributed AI workloads.
NVIDIA AI Software Stack
Access NVIDIA NIM microservices, AI frameworks, enterprise libraries and deployment tools.
Maximise Your DGX Spark With Production Ready Software
NVIDIA AI Enterprise accelerates AI development using pre-configured frameworks, NVIDIA NIM and NeMo microservices, reducing software complexity and improving deployment speed.
Learn More
NVIDIA DGX Spark Workloads
Develop AI Models
Prototype, train and deploy advanced AI workloads locally or across cloud and datacenter infrastructure.
Fine Tune LLMs
Fine-tune modern reasoning and generative AI models using NVIDIA accelerated unified memory systems.
Robot AI Models
Build robotics and machine vision applications with NVIDIA Isaac and edge AI frameworks.
Data Science
Accelerate analytics, machine learning and AI pipelines with NVIDIA enterprise compute technologies.
Frequently Asked Questions
What is NVIDIA DGX Spark used for?
NVIDIA DGX Spark is designed as a personal AI workstation for developers, researchers, data scientists and enterprise AI teams.
Can DGX Spark run large language models?
Yes. DGX Spark is optimised for training and fine-tuning modern reasoning and generative AI models efficiently.
What are the specifications of DGX Spark?
DGX Spark features Blackwell GPU architecture, unified memory, high-speed networking and enterprise AI software stack support.