PNY DGX Spark Desktop AI Computer with NVIDIA Grace Blackwell Architecture - 128GB RAM, 4TB SSD, Compact Form Factor

PNY DGX Spark Desktop AI Computer - ARM Cortex X925 - 128 GB - 4 TB SSD - Compact

The PNY DGX Spark Desktop AI Computer is a powerful, compact solution designed for advanced AI development and deployment. Featuring NVIDIA's innovative Grace Blackwell architecture, this system combines a 20-core Arm CPU with 10 Cortex-X925 and 10 Cortex-A725 cores, delivering exceptional processing capabilities for AI workloads. With 1,000 AI TOPS of tensor performance, it handles large models with up to 200 billion parameters seamlessly, making it ideal for research, development, and deployment in data centers or cloud environments. The system includes 128GB of LPDDR5x unified memory, providing high bandwidth and fast data access, critical for high-performance AI tasks. Equipped with Gen 5 Tensor Cores and Gen 4RT Cores, the DGX Spark ensures accelerated AI training and inference, optimizing performance for complex neural network models.

Designed for versatility, the DGX Spark features a range of connectivity options including four USB Type-C ports, 10 GbE Ethernet, Wi-Fi 7, and Bluetooth 5.3, enabling effortless integration into existing infrastructure. Its rich multimedia support includes HDMI 2.1a for multichannel audio output, while NVENC and NVDEC hardware accelerate video encoding and decoding processes. Running on DGX OS, the system ensures a stable and optimized environment for AI applications. Weighing approximately 2.65 pounds and housed in a compact form factor, this AI computer brings high-end computational power into a space-efficient design suitable for desktop deployment. Backed by a one-year limited warranty, the PNY DGX Spark is built to meet the demanding needs of AI researchers and developers.

With its combination of cutting-edge hardware and versatile connectivity, the PNY DGX Spark simplifies AI model development from prototyping to deployment. The inclusion of the ConnectX-7 Smart NIC enhances network performance, enabling high-speed data transfer essential for large-scale AI operations. This system's robust architecture and extensive feature set make it an excellent choice for organizations seeking a compact yet powerful AI solution that integrates seamlessly into existing workflows and accelerates innovation.

Features:

• NVIDIA Grace Blackwell architecture with 20-core Arm CPU, including 10 Cortex-X925 and 10 Cortex-A725 cores, for high-efficiency processing.

• Gen 5 Tensor Cores and Gen 4RT Cores provide accelerated AI training and inference capabilities.

• Tensor performance reaches up to 1,000 AI TOPS, supported by 128GB LPDDR5x unified memory and a 256-bit memory interface with 273 GB/s bandwidth.

• Comprehensive connectivity options including 4 USB Type-C ports, 10 GbE Ethernet, Wi-Fi 7, and Bluetooth 5.3 facilitate seamless integration.

• Multichannel audio output via HDMI 2.1a, hardware-accelerated video encoding/decoding with NVENC and NVDEC, and a compact, lightweight design ideal for desktop AI deployment.