Skip to content

MikroTik WiFi 7 Giveawy by Baltic Networks

In-stock gear ready to ship • Free shipping on orders $100+
In-stock gear ready to ship • Free shipping on orders $150+
NVIDIA DGX Spark AI Supercomputer

NVIDIA DGX Spark AI Supercomputer

by NVIDIA
SKU 940-54242-0000-000
$7,414.00 - $7,414.00
$7,414.00
$7,414.00
$7,414.00 - $7,414.00
$7,414.00

Choose Your DGX Spark Setup

Single DGX Spark 2 Sparks + Cable AI Cluster Bundle

Product Information

Overview

The NVIDIA DGX Spark AI Supercomputer is a personal AI computer designed for developing, testing, and running advanced AI models directly from your desktop. Powered by the NVIDIA GB10 Grace Blackwell Superchip, it combines AI compute, high-speed memory, and enterprise-class software into a compact desktop platform.

Built for AI developers, researchers, data scientists, and software engineers, DGX Spark delivers up to 1 petaFLOP of AI performance using FP4 precision and supports AI models with up to 200 billion parameters. NVIDIA ConnectX networking allows two DGX Spark systems to work together with models up to 405 billion parameters, while NVIDIA DGX OS, Ubuntu Linux, and the NVIDIA AI software stack provide everything needed to begin developing AI applications locally.

Ideal Applications

  • AI model development and prototyping
  • Fine-tuning and inference for large language models
  • Building autonomous AI agents
  • Computer vision and speech AI applications
  • Local AI research and experimentation
  • Desktop AI development before cloud deployment

Highlights

Key Features

Built for Local AI Development

Develop, test, fine-tune, and run advanced AI workloads directly from your desktop using the NVIDIA GB10 Grace Blackwell Superchip and the complete NVIDIA AI software stack.

Supports Massive AI Models

Run AI models with up to 200 billion parameters on a single system, or connect two DGX Spark systems using NVIDIA ConnectX networking to support models with up to 405 billion parameters.

High-Performance Unified Memory

128 GB of coherent unified system memory provides the capacity and bandwidth needed for modern AI development, helping reduce bottlenecks when working with large models.

Desktop to Data Center Workflow

Prototype locally using NVIDIA DGX OS and Ubuntu Linux, then move workloads to NVIDIA DGX Cloud or accelerated data center infrastructure using the same software platform.

Product Details

Technical Specifications

System Specifications

  • Architecture: NVIDIA Grace Blackwell
  • GPU: NVIDIA Blackwell Architecture
  • CPU: 20-core Arm (10 Cortex-X925 + 10 Cortex-A725)
  • CUDA Cores: NVIDIA Blackwell Generation
  • Tensor Cores: 5th Generation
  • RT Cores: 4th Generation
  • AI Performance: Up to 1 PFLOP (FP4)

Memory & Storage

  • System Memory: 128 GB LPDDR5x coherent unified memory
  • Memory Interface: 256-bit
  • Memory Bandwidth: Up to 273 GB/s
  • Storage: 4 TB NVMe M.2 with self-encryption

Networking & Connectivity

  • Ethernet: 1 × 10 GbE RJ-45
  • NIC: NVIDIA ConnectX-7 @ 200 Gbps
  • Wi-Fi: Wi-Fi 7
  • Bluetooth: Bluetooth 5.4 with LE
  • USB: 4 × USB Type-C
  • Display Output: 1 × HDMI 2.1a
  • Audio: HDMI multichannel audio output

Software

  • Operating System: NVIDIA DGX OS

Physical

  • Dimensions: 150 × 150 × 50.5 mm
  • Weight: 1.2 kg
  • Power Consumption: 240 W

Resources

Product Documentation

Datasheet