PNY NVIDIA DGX Spark Founder’s Edition

Compact Powerhouse for Next-Generation AI Innovation

At Promise Computer Technology LLC, we are committed to delivering cutting-edge AI infrastructure that empowers organizations to innovate faster and smarter. The PNY NVIDIA DGX Spark Founder’s Edition represents a breakthrough in compact AI computing—bringing data-centre–class performance to a remarkably small and efficient form factor.

Built on the revolutionary NVIDIA Grace Blackwell architecture, DGX Spark is designed for developers, researchers and enterprises seeking unmatched AI performance, scalability and efficiency at the edge or in dedicated AI environments.

Built on NVIDIA Grace Blackwell Architecture

The DGX Spark Founder’s Edition is powered by the NVIDIA GB10 Grace Blackwell Superchip, a tightly integrated CPU–GPU platform engineered specifically for AI workloads. This advanced architecture delivers up to 1 petaflop of AI performance (FP4), enabling rapid training, fine-tuning and inference for large language models, computer vision and advanced analytics.

High-Performance Compute in a Compact Design

Despite its ultra-compact footprint, DGX Spark offers enterprise-grade compute capabilities:

• GPU: NVIDIA GB10 Grace Blackwell Superchip
• CPU: 20-core Arm processor
• 10 × Cortex-X925 (high-performance cores)
• 10 × Cortex-A725 (efficiency cores)

This balanced CPU–GPU integration ensures optimal performance across AI, data processing, and orchestration workloads.

Unified Memory for AI Acceleration

The system features 128 GB of LPDDR5x unified system memory, allowing the CPU and GPU to access a shared memory pool. This design dramatically reduces data movement, improves latency, and accelerates AI pipelines—especially for memory-intensive models and real-time applications.

Secure, High-Speed Storage

DGX Spark comes equipped with a 4 TB NVMe M.2 self-encrypting SSD, delivering:

Lightning-fast data access for AI datasets

Enhanced security for sensitive enterprise and research data

Reliable local storage for model training and inference

NVIDIA DGX OS – AI-Optimized Software Stack

The system runs NVIDIA DGX OS, a Linux-based operating environment optimized for AI workloads. DGX OS provides seamless integration with NVIDIA’s AI ecosystem, including CUDA, cuDNN, TensorRT and popular AI frameworks—ensuring faster deployment and simplified management.

Advanced Connectivity and Networking

DGX Spark is designed for both standalone performance and scalable multi-node AI deployments:

Small Footprint, Big Impact

With dimensions of just 150 mm × 150 mm × 50.5 mm and a weight of 1.2 kg, DGX Spark delivers exceptional performance in a space-saving design—ideal for AI labs, edge deployments and enterprise environments where space and power efficiency matter.

Reliability and Support

The PNY NVIDIA DGX Spark Founder’s Edition is backed by a 1-year limited warranty, ensuring peace of mind and dependable operation for mission-critical AI workloads.