{"product_id":"lenovo-nvidia-gb300-nvl72-server-ai-platform","title":"Lenovo NVIDIA GB300 NVL72 | Server | AI Platform","description":"\u003cstyle\u003e\na {\n    text-decoration: none;\n    color: #464feb;\n}\ntr th, tr td {\n    border: 1px solid #e6e6e6;\n}\ntr th {\n    background-color: #f5f5f5;\n}\n\u003c\/style\u003e\n\u003cdiv\u003e\n\u003cp\u003e\u003cstrong\u003eLenovo NVIDIA GB300 NVL72 AI Platform\u003c\/strong\u003e\u003cbr\u003e\u003cstrong\u003e72 NVIDIA Blackwell GPUs | NVIDIA Grace CPUs | Rack-Scale Enterprise AI Platform\u003c\/strong\u003e\u003c\/p\u003e\n\u003cp\u003eThe \u003cstrong\u003eLenovo NVIDIA GB300 NVL72\u003c\/strong\u003e is a rack-scale AI platform engineered to accelerate large-scale AI training, inference, reasoning, and high-performance computing workloads. Built on the powerful \u003cstrong\u003eNVIDIA Grace-Blackwell architecture\u003c\/strong\u003e, the system combines NVIDIA Grace CPUs and NVIDIA Blackwell GPUs with fifth-generation NVIDIA NVLink technology to create a unified AI infrastructure designed for enterprise-scale deployments.\u003cmeta content=\"text\/html; charset=us-ascii\" http-equiv=\"Content-Type\"\u003e\u003c\/p\u003e\n\u003c\/div\u003e\n\u003cstyle\u003e\na {\n    text-decoration: none;\n    color: #464feb;\n}\ntr th, tr td {\n    border: 1px solid #e6e6e6;\n}\ntr th {\n    background-color: #f5f5f5;\n}\n\u003c\/style\u003e\n\u003cdiv\u003e\n\u003cdiv\u003eDesigned for modern AI factories and enterprise AI environments, the platform delivers massive GPU acceleration across a unified 72-GPU domain, making it ideal for large language models (LLMs), generative AI, agentic AI, vector databases, and advanced AI inference workloads. High-speed networking, extensive memory bandwidth, and rack-scale NVLink connectivity provide the performance required for demanding AI and HPC applications.\u003c\/div\u003e\n\u003cdiv\u003e\n\u003cmeta content=\"text\/html; charset=us-ascii\" http-equiv=\"Content-Type\"\u003e\n\u003cbr\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003cstyle\u003e\na {\n    text-decoration: none;\n    color: #464feb;\n}\ntr th, tr td {\n    border: 1px solid #e6e6e6;\n}\ntr th {\n    background-color: #f5f5f5;\n}\n\u003c\/style\u003e\n\u003cdiv\u003e\n\u003cdiv\u003e\n\u003cdiv\u003eThe Lenovo NVIDIA GB300 NVL72 features an advanced hybrid liquid-air cooling architecture that removes heat directly at the source, enabling exceptional compute density while maintaining predictable performance and energy efficiency. Warm-water operation further improves cooling efficiency and reduces operational costs, making it an ideal platform for next-generation data centers.\u003c\/div\u003e\n\u003cdiv\u003e\n\u003cmeta content=\"text\/html; charset=us-ascii\" http-equiv=\"Content-Type\"\u003e\n\u003cbr\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003cstyle\u003e\na {\n    text-decoration: none;\n    color: #464feb;\n}\ntr th, tr td {\n    border: 1px solid #e6e6e6;\n}\ntr th {\n    background-color: #f5f5f5;\n}\n\u003c\/style\u003e\n\u003cdiv\u003e\n\u003cdiv\u003e\n\u003cdiv\u003e\n\u003cdiv\u003e\n\u003cp\u003e\u003cstrong\u003eKey Features\u003c\/strong\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003e\u003cstrong\u003eNVIDIA Grace-Blackwell architecture optimized for enterprise AI and HPC\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eUnified 72-GPU rack-scale NVLink domain for large-scale AI workloads\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003e5th-generation NVIDIA NVLink technology for coherent CPU-GPU memory access\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003e72 NVIDIA Blackwell GPUs interconnected across the rack\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003e2 NVIDIA Grace CPUs and 4 NVIDIA Blackwell GPUs per compute tray\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003e18 compute trays within a 48U MGX rack architecture\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003e9 NVLink switch trays for rack-scale GPU interconnectivity\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eHBM3e GPU memory architecture optimized for AI training and inference\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eLPDDR5X CPU memory for high-bandwidth data processing\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eIntegrated 800Gb\/s CX8 networking for high-performance communications\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eSupport for InfiniBand or Ethernet scale-out clustering\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eHybrid liquid-air cooling with warm-water operation from 18°C to 45°C\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eLiquid cooling applied to CPUs, GPUs, voltage regulators, and NVLink modules\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eUp to 8 x 33kW power shelves with balanced 48V busbar distribution\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eNVIDIA OpenBMC-based systems management\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eDesigned for LLM training, AI inference, vector databases, and agentic AI workloads\u003c\/strong\u003e\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003e3-Year Onsite Next Business Day Warranty\u003c\/strong\u003e\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cstrong\u003eSpecifications\u003c\/strong\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003eForm Factor: 1U Grace-Blackwell Compute Tray\u003c\/li\u003e\n\u003cli\u003e18 Compute Trays in a 48U MGX Rack\u003c\/li\u003e\n\u003cli\u003e9 x 1U NVLink Switch Trays\u003c\/li\u003e\n\u003cli\u003e2 x NVIDIA Grace CPUs per Compute Tray\u003c\/li\u003e\n\u003cli\u003e4 x NVIDIA Blackwell GPUs per Compute Tray\u003c\/li\u003e\n\u003cli\u003e72 NVIDIA Blackwell GPUs per Rack\u003c\/li\u003e\n\u003cli\u003e5th-Generation NVIDIA NVLink Technology\u003c\/li\u003e\n\u003cli\u003eHBM3e GPU Memory Architecture\u003c\/li\u003e\n\u003cli\u003eHigh-Bandwidth LPDDR5X CPU Memory\u003c\/li\u003e\n\u003cli\u003eUp to 8 x E1.S NVMe SSDs per Compute Tray\u003c\/li\u003e\n\u003cli\u003eIntegrated 800Gb\/s CX8 Networking\u003c\/li\u003e\n\u003cli\u003eDual 25GbE LOM Connectivity\u003c\/li\u003e\n\u003cli\u003eRack-Scale NVLink Domain Across 72 GPUs\u003c\/li\u003e\n\u003cli\u003eSupport for High-Speed InfiniBand or Ethernet Scale-Out Clusters\u003c\/li\u003e\n\u003cli\u003eHybrid Liquid + Air Cooling System\u003c\/li\u003e\n\u003cli\u003eSupports PG25 or Deionized Water Cooling\u003c\/li\u003e\n\u003cli\u003eWarm Water Operation: 18°C to 45°C\u003c\/li\u003e\n\u003cli\u003eUp to 8 x 33kW Power Shelves\u003c\/li\u003e\n\u003cli\u003eBalanced 48V Busbar Power Distribution\u003c\/li\u003e\n\u003cli\u003eAdvanced Monitoring via Confluent and NVIDIA Tools\u003c\/li\u003e\n\u003cli\u003eNVIDIA OpenBMC-Based Systems Management\u003c\/li\u003e\n\u003cli\u003eSupported Operating System: Ubuntu 22.04 (HWE)\u003c\/li\u003e\n\u003cli\u003eStandard Warranty: 3-Year Onsite Next Business Day Service\u003c\/li\u003e\n\u003cli\u003eOptional Warranty: 5-Year Service Available\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cstrong\u003eIdeal For\u003c\/strong\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003eLarge Language Model (LLM) Training\u003c\/li\u003e\n\u003cli\u003eGenerative AI Platforms\u003c\/li\u003e\n\u003cli\u003eAI Reasoning and Agentic AI\u003c\/li\u003e\n\u003cli\u003eEnterprise AI Infrastructure\u003c\/li\u003e\n\u003cli\u003eAI Inference at Scale\u003c\/li\u003e\n\u003cli\u003eVector Databases\u003c\/li\u003e\n\u003cli\u003eDeep Learning Workloads\u003c\/li\u003e\n\u003cli\u003eHigh Performance Computing (HPC)\u003c\/li\u003e\n\u003cli\u003eScientific Computing and Simulations\u003c\/li\u003e\n\u003cli\u003eLarge-Scale AI Factories\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e","brand":"Eprom Data Systems","offers":[{"title":"Default Title","offer_id":51447182164135,"sku":"Lenovo NVIDIA GB300 NVL72","price":399999.0,"currency_code":"SGD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0697\/2735\/5047\/files\/Picture44_08cfff46-c2d4-4d2c-8815-a10c7152d6c4.png?v=1790838263","url":"https:\/\/eprom.com.sg\/products\/lenovo-nvidia-gb300-nvl72-server-ai-platform","provider":"Eprom Data Systems","version":"1.0","type":"link"}