{"product_id":"nvidia-dgx-spark-ai-supercomputer-gb10-grace-blackwell-superchip-20-core-arm-cpu-128gb-lpddr5x-unified-memory-4tb-nvme-m-2-ssd-1-pflop-fp4-ai-performance","title":"NVIDIA DGX Spark AI Supercomputer | GB10 Grace Blackwell Superchip | 20-Core ARM CPU | 128GB LPDDR5x Unified Memory | 4TB NVMe M.2 SSD | 1 PFLOP FP4 AI Performance","description":"\u003cdiv style=\"text-align: justify; font-family: Arial, sans-serif; font-size: 15px; line-height: 1.7; color: #333;\"\u003e\n\u003ch2\u003eNVIDIA DGX Spark AI Supercomputer | GB10 Grace Blackwell Superchip | 20-Core ARM CPU | 128GB LPDDR5x Unified Memory | 4TB NVMe M.2 SSD | 1 PFLOP FP4 AI Performance\u003c\/h2\u003e\n\u003cp\u003eThe \u003cstrong\u003eNVIDIA DGX Spark AI Supercomputer\u003c\/strong\u003e is a compact desktop AI computing system designed for developers, researchers, data scientists, and AI professionals who need powerful local AI development capabilities. Powered by the NVIDIA GB10 Grace Blackwell Superchip, it combines a 20-core Arm CPU with an NVIDIA Blackwell GPU, delivering advanced computing performance for AI model development, inference, fine-tuning, and deployment. Configured with \u003cstrong\u003e128GB LPDDR5x unified system memory\u003c\/strong\u003e and a \u003cstrong\u003e4TB NVMe M.2 SSD\u003c\/strong\u003e, NVIDIA DGX Spark provides substantial memory capacity and high-speed storage for demanding AI workloads. Its integrated NVIDIA AI software stack and compact desktop design make it suitable for building and testing AI applications locally, with workflows that can also scale to data centre and cloud environments.\u003c\/p\u003e\n\u003cp\u003e\u003cimg src=\"https:\/\/cdn.shopify.com\/s\/files\/1\/0898\/6224\/2671\/files\/6_f4e2691d-5042-4323-9f7e-9ce8b28084d3.jpg?v=1793823711\" alt=\"\"\u003e\u003cimg src=\"https:\/\/cdn.shopify.com\/s\/files\/1\/0898\/6224\/2671\/files\/7_356f0b76-3015-4240-b392-7a923e4062b8.jpg?v=1793823711\" alt=\"\"\u003e\u003cimg src=\"https:\/\/cdn.shopify.com\/s\/files\/1\/0898\/6224\/2671\/files\/8_94d3308a-f50b-442d-bd1e-61129ca4c6bd.jpg?v=1793823711\" alt=\"\"\u003e\u003c\/p\u003e\n\u003ch3\u003eKey Features\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA GB10 Grace Blackwell Superchip:\u003c\/strong\u003e Combines CPU and GPU computing capabilities in a unified architecture designed for advanced AI workloads.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e20-Core Arm CPU:\u003c\/strong\u003e Features 10 Cortex-X925 and 10 Cortex-A725 cores for efficient processing and multitasking.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA Blackwell GPU Architecture:\u003c\/strong\u003e Includes 5th-generation Tensor Cores and 4th-generation RT Cores for accelerated AI and graphics-related workloads.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eUp to 1 PFLOP FP4 AI Performance:\u003c\/strong\u003e Delivers theoretical FP4 AI computing performance using the sparsity feature.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e128GB LPDDR5x Unified Memory:\u003c\/strong\u003e Provides high-capacity coherent memory shared across the integrated CPU and GPU architecture.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e4TB NVMe M.2 SSD:\u003c\/strong\u003e Provides high-speed storage capacity for AI applications, datasets, models, and development projects.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA ConnectX-7 SmartNIC:\u003c\/strong\u003e Supports high-speed networking with ConnectX-7 technology.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eWi-Fi 7 and 10GbE:\u003c\/strong\u003e Provides advanced wireless and wired network connectivity for professional environments.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eCompact Desktop Design:\u003c\/strong\u003e Delivers powerful AI computing in a small-form-factor system suitable for desktop environments.\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ch3\u003eTechnical Specifications\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eProduct Model:\u003c\/strong\u003e NVIDIA DGX Spark\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eAI Computing Architecture:\u003c\/strong\u003e NVIDIA Grace Blackwell\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eSuperchip:\u003c\/strong\u003e NVIDIA GB10 Grace Blackwell Superchip\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eCPU:\u003c\/strong\u003e 20-Core Arm Processor\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eCPU Core Configuration:\u003c\/strong\u003e 10 Cortex-X925 + 10 Cortex-A725\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eGPU Architecture:\u003c\/strong\u003e NVIDIA Blackwell\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eTensor Cores:\u003c\/strong\u003e 5th Generation\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eRT Cores:\u003c\/strong\u003e 4th Generation\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eAI Performance:\u003c\/strong\u003e Up to 1 PFLOP FP4 (theoretical performance using sparsity)\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eSystem Memory:\u003c\/strong\u003e 128GB LPDDR5x Coherent Unified Memory\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eMemory Interface:\u003c\/strong\u003e 256-bit\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eMemory Bandwidth:\u003c\/strong\u003e Up to 273GB\/s\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eStorage Capacity:\u003c\/strong\u003e 4TB NVMe M.2 SSD\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNetwork Interface:\u003c\/strong\u003e NVIDIA ConnectX-7 SmartNIC\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eEthernet:\u003c\/strong\u003e 1 × RJ-45 10GbE\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eWireless Connectivity:\u003c\/strong\u003e Wi-Fi 7\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eBluetooth:\u003c\/strong\u003e Bluetooth 5.4\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eUSB Connectivity:\u003c\/strong\u003e 4 × USB Type-C\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eDisplay Output:\u003c\/strong\u003e 1 × HDMI 2.1a; DisplayPort over USB-C with DisplayPort Alternate Mode\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eVideo Processing:\u003c\/strong\u003e 1 × NVENC and 1 × NVDEC\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eOperating System:\u003c\/strong\u003e NVIDIA DGX OS\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003ePower Adapter:\u003c\/strong\u003e 240W\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eGB10 TDP:\u003c\/strong\u003e 140W\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eDimensions:\u003c\/strong\u003e 150 × 150 × 50.5mm\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eWeight:\u003c\/strong\u003e Approximately 1.2kg\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ch3\u003eAdvanced AI Computing with NVIDIA Grace Blackwell\u003c\/h3\u003e\n\u003cp\u003eThe NVIDIA GB10 Grace Blackwell Superchip brings together a powerful Arm-based CPU and NVIDIA Blackwell GPU architecture. This integrated design is developed to accelerate AI workloads, allowing professionals to work with modern AI frameworks and applications in a compact desktop environment. With 5th-generation Tensor Cores and advanced AI computing capabilities, DGX Spark is suitable for developing, testing, and running demanding AI models.\u003c\/p\u003e\n\u003ch3\u003e128GB Unified Memory and 4TB High-Speed Storage\u003c\/h3\u003e\n\u003cp\u003eWith 128GB of LPDDR5x coherent unified memory, NVIDIA DGX Spark provides a large shared memory pool for CPU and GPU workloads. The 256-bit memory interface and bandwidth of up to 273GB\/s support data-intensive AI processing. Its \u003cstrong\u003e4TB NVMe M.2 SSD\u003c\/strong\u003e provides high-speed storage capacity for AI models, large datasets, software environments, and development projects.\u003c\/p\u003e\n\u003ch3\u003eHigh-Performance Networking and Connectivity\u003c\/h3\u003e\n\u003cp\u003eDGX Spark includes NVIDIA ConnectX-7 SmartNIC technology, 10GbE wired networking, and Wi-Fi 7 wireless connectivity. These interfaces support fast data movement and network integration in professional development environments. USB Type-C connectivity and HDMI 2.1a provide options for connecting compatible peripherals and displays.\u003c\/p\u003e\n\u003ch3\u003eCompact Desktop AI Supercomputer\u003c\/h3\u003e\n\u003cp\u003eDesigned for desktop deployment, NVIDIA DGX Spark combines substantial AI computing capabilities in a compact chassis measuring approximately 150 × 150 × 50.5mm. Its small footprint makes it suitable for professional workspaces, development labs, research environments, and offices where dedicated AI computing is required without a traditional large server installation.\u003c\/p\u003e\n\u003ch3\u003eIdeal Applications\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003eArtificial Intelligence and Machine Learning Development\u003c\/li\u003e\n\u003cli\u003eLarge Language Model (LLM) Development and Inference\u003c\/li\u003e\n\u003cli\u003eAI Model Prototyping and Fine-Tuning\u003c\/li\u003e\n\u003cli\u003eGenerative AI Application Development\u003c\/li\u003e\n\u003cli\u003eData Science and Advanced Computing Research\u003c\/li\u003e\n\u003cli\u003eAI Agent Development and Testing\u003c\/li\u003e\n\u003cli\u003eComputer Vision and Deep Learning Projects\u003c\/li\u003e\n\u003cli\u003eLocal AI Workloads and Professional Development Environments\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ch3\u003eBenefits of NVIDIA DGX Spark\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003ePowerful Local AI Computing:\u003c\/strong\u003e Develop and run AI workloads locally using NVIDIA Grace Blackwell architecture.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eLarge Unified Memory:\u003c\/strong\u003e The 128GB memory configuration supports demanding workloads that require substantial shared system memory.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eHigh-Capacity Storage:\u003c\/strong\u003e The 4TB NVMe SSD provides extensive space for AI models, datasets, applications, and development environments.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eAdvanced AI Acceleration:\u003c\/strong\u003e Blackwell GPU architecture and Tensor Cores support accelerated AI processing.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eProfessional Networking:\u003c\/strong\u003e ConnectX-7, 10GbE, and Wi-Fi 7 provide advanced connectivity options.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eSpace-Saving Design:\u003c\/strong\u003e Its compact form factor fits easily into professional desktop and laboratory environments.\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ch3\u003eWhy Choose the NVIDIA DGX Spark?\u003c\/h3\u003e\n\u003cp\u003eNVIDIA DGX Spark is designed for professionals who need a dedicated AI development platform with NVIDIA's integrated hardware and software ecosystem. Its GB10 Grace Blackwell Superchip, 128GB unified memory, 4TB NVMe storage, and advanced networking make it a capable solution for local AI development, model experimentation, and professional machine learning workflows. It offers a compact alternative to deploying larger AI infrastructure for supported development and inference workloads.\u003c\/p\u003e\n\u003ch3\u003eCompatibility and Important Considerations\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003eDesigned for AI development, inference, prototyping, and fine-tuning workloads using supported NVIDIA software and frameworks.\u003c\/li\u003e\n\u003cli\u003eAI model capacity and performance depend on model architecture, precision, software configuration, and workload requirements.\u003c\/li\u003e\n\u003cli\u003eThe advertised 1 PFLOP FP4 performance is a theoretical figure using the sparsity feature and should not be interpreted as guaranteed performance for every application.\u003c\/li\u003e\n\u003cli\u003eUse the appropriate supplied power adapter for the intended system operation.\u003c\/li\u003e\n\u003cli\u003eCheck software, peripheral, display, and network compatibility before deployment in an existing professional environment.\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ch3\u003eWhy Buy from IT Splendid (UAE)?\u003c\/h3\u003e\n\u003cp\u003eIT Splendid offers advanced computing hardware, professional workstations, networking equipment, and enterprise technology solutions for individuals and businesses across the UAE. The \u003cstrong\u003eNVIDIA DGX Spark AI Supercomputer\u003c\/strong\u003e is an advanced AI computing solution for developers, researchers, and organisations looking to build and operate local artificial intelligence applications using NVIDIA Grace Blackwell technology.\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eAdvanced Computing Solutions:\u003c\/strong\u003e Explore desktop computers, workstations, servers, AI hardware, and professional computing equipment.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eEnterprise Technology Selection:\u003c\/strong\u003e Discover networking products, business hardware, and technology solutions for professional environments.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eUAE Delivery Options:\u003c\/strong\u003e Delivery services are available across the UAE, subject to current store terms.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eConvenient Online Shopping:\u003c\/strong\u003e Browse product specifications and order through the IT Splendid online store.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eFlexible Payment Methods:\u003c\/strong\u003e Choose from available payment options at checkout, including Tabby and Tamara where eligible.\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003eUpgrade your artificial intelligence development environment with the \u003cstrong\u003eNVIDIA DGX Spark AI Supercomputer\u003c\/strong\u003e. Featuring the NVIDIA GB10 Grace Blackwell Superchip, 20-core ARM CPU, NVIDIA Blackwell GPU, 128GB unified memory, and \u003cstrong\u003e4TB NVMe M.2 SSD\u003c\/strong\u003e, it provides a powerful compact computing platform for AI developers, researchers, data scientists, and businesses across the UAE.\u003c\/p\u003e\n\u003c\/div\u003e","brand":"NVIDIA","offers":[{"title":"Default Title","offer_id":68022876143983,"sku":"DGX-Spark-4TB-QUOTE","price":0.0,"currency_code":"AED","in_stock":false}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0898\/6224\/2671\/files\/1_3d4ce7a6-3451-4b0f-b6f3-5b2e7b180b07.jpg?v=1793823711","url":"https:\/\/www.itsplendid.com\/products\/nvidia-dgx-spark-ai-supercomputer-gb10-grace-blackwell-superchip-20-core-arm-cpu-128gb-lpddr5x-unified-memory-4tb-nvme-m-2-ssd-1-pflop-fp4-ai-performance","provider":"IT Splendid","version":"1.0","type":"link"}