NVIDIA DGX Spark, 20 core Arm, 10 Cortex-X925 + Cortex-A725 Arm; 128GB LPDDR5x Unified RAM; 4TB NVMe M2 SSD Blackwell Architecture, 10GbE, WiFi 7, Bluetooth 5.3
Featured
|
£ 4,899.99 |
To Shop
|
Description
Amazon
Unmatched AI Performance: Powered by the NVIDIA GB10 Grace Blackwell Superchip, delivering 1 petaFLOP of AI performance for cutting-edge model processing. Advanced Processing Power: Features a 20-core Arm CPU (10 Cortex-X925 + 10 Cortex-A725) for seamless handling of complex AI workloads. High-Capacity Memory: Equipped with 128GB LPDDR5x Unified RAM to support large-scale AI model prototyping and fine-tuning. Massive Storage: Includes a 4TB NVMe M.2 SSD for rapid data access and ample storage for AI datasets and models. Next-Gen Connectivity: Supports WiFi 7, 10GbE LAN, and Bluetooth 5.3 for fast, reliable networking and device integration. Compact & Portable Design: Sleek horizontal form factor (5.91" x 5.91" x 1.99", 2.65 lbs) for easy deployment in any environment. Ready for AI Innovation: Preinstalled NVIDIA DGX OS and AI software stack enable local and cloud-based deployment of models up to 200 billion parameters.
Compare webshops (1)
Unmatched AI Performance: Powered by the NVIDIA GB10 Grace Blackwell Superchip, delivering 1 petaFLOP of AI performance for cutting-edge model processing. Advanced Processing Power: Features a 20-core Arm CPU (10 Cortex-X925 + 10 Cortex-A725) for seamless handling of complex AI workloads. High-Capacity Memory: Equipped with 128GB LPDDR5x Unified RAM to support large-scale AI model prototyping and fine-tuning. Massive Storage: Includes a 4TB NVMe M.2 SSD for rapid data access and ample storage for AI datasets and models. Next-Gen Connectivity: Supports WiFi 7, 10GbE LAN, and Bluetooth 5.3 for fast, reliable networking and device integration. Compact & Portable Design: Sleek horizontal form factor (5.91" x 5.91" x 1.99", 2.65 lbs) for easy deployment in any environment. Ready for AI Innovation: Preinstalled NVIDIA DGX OS and AI software stack enable local and cloud-based deployment of models up to 200 billion parameters.