{"product_id":"nvidia-dgx-b300-ai-server-8x-blackwell-ultra-gpu","title":"NVIDIA DGX B300 AI Server 8x Blackwell Ultra GPU","description":"\u003cp style=\"margin:1em 0\"\u003eThe NVIDIA DGX B300 is a fully integrated AI factory server built on the NVIDIA Blackwell Ultra architecture, combining eight Blackwell Ultra SXM GPUs with 2.1 TB of total GPU memory and 14.4 TB\/s of aggregate NVLink bandwidth. With up to 144 PFLOPS of FP4 inference and 72 PFLOPS of FP8 training performance, it is engineered to train trillion-parameter models and deploy real-time reasoning engines while handling massive datasets without memory bottlenecks. Its 8x NVIDIA ConnectX-8 VPI adapters deliver up to 800 Gb\/s per port, enabling seamless scaling across NVIDIA DGX SuperPOD and BasePOD clusters. The system ships with NVIDIA AI Enterprise, NVIDIA DGX OS, and NVIDIA Mission Control, providing a pre-optimized software stack for generative AI, data science, and computer vision. Available in an air-cooled 8U configuration, it pairs dual Intel Xeon 6776P processors with 2 TB of system memory, configurable to 4 TB, and internal storage of eight 3.84 TB NVMe E1.S drives.\u003c\/p\u003e\n\u003ch3 style=\"margin:1.25em 0 0.5em\"\u003eKey Features\u003c\/h3\u003e\n\u003ctable class=\"pu-table\" style=\"width:100%;border-collapse:separate;border-spacing:0;border:0;box-shadow:none;outline:none;background:none;table-layout:fixed;margin:0 0 1em\"\u003e\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eBrand\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eNVIDIA\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eModel\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eDGX B300\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eProduct Type\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eAI Server\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eGPU\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e8x NVIDIA Blackwell Ultra SXM\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eTotal GPU Memory\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e2.1 TB total, 62 TB\/s HBM3e bandwidth\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003ePerformance\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eFP4 Tensor Core* - 144 PFLOPS | 108 PFLOPS; FP8\/FP6 Tensor Core** - 72 PFLOPS\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eSystem Memory\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e2 TB, configurable to 4 TB\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eNVIDIA NVLink Switch System\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e2x\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eNVIDIA NVLink Bandwidth\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e14.4 TB\/s aggregate bandwidth\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eSystem Power\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e14.5 kW (Busbar); 15.1 kW (PSU)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eCPU\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eIntel Xeon 6776P processors\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eNetworking\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e8x OSFP ports serving 8x NVIDIA ConnectX-8 VPI \u0026gt;Up to 800 Gb\/s of NVIDIA InfiniBand\/Ethernet; 2x dual-port QSFP112 NVIDIA BlueField®-3 DPU \u0026gt;Up to 400 Gb\/s of NVIDIA InfiniBand\/Ethernet\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eManagement Network\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e1GbE onboard network interface card (NIC) with RJ45; 1GbE RJ45 host baseboard management controller (BMC)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eStorage\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eOS: 2x 1.9 TB NVMe M.2; Internal storage: 8x 3.84 TB NVMe E1.S\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eSoftware\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eNVIDIA AI Enterprise (optimized AI software); NVIDIA Mission Control (AI data center operations and orchestration with NVIDIA Run:ai technology); NVIDIA DGX OS (operating system); Supports Red Hat Enterprise Linux \/ Rocky \/ Ubuntu\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eRack Units\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e10\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eOperating Temperature\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e50°F–95°F\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eProduct Dimensions\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e35.3 in L x 19 in W x 17.5 in H\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eSupport\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003eThree-year business-standard hardware and software support\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd style=\"border:0;padding:0.4em 0.75em 0.4em 0;opacity:0.65;vertical-align:top\"\u003eCondition\u003c\/td\u003e\n\u003ctd style=\"border:0;padding:0.4em 0;vertical-align:top\"\u003eNew, Factory-Sealed\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\u003c\/table\u003e\n\u003ch3 style=\"margin:1.25em 0 0.5em\"\u003eHighlights\u003c\/h3\u003e\n\u003cp style=\"margin:1em 0\"\u003eDelivering a massive 2.3TB of GPU memory and 144 petaFLOPS of inference performance, this turnkey system is purpose-built to accelerate massive LLM training, complex reasoning, and data-intensive workloads. Designed for the modern data center, the DGX B300 offers 11x faster inference and 4x faster training than previous generations, enabling enterprises to refine, deploy, and scale AI with hyperscale efficiency.\u003c\/p\u003e\n\u003ch3 style=\"margin:1.25em 0 0.5em\"\u003eCompatibility\u003c\/h3\u003e\n\u003cp style=\"margin:1em 0\"\u003eSupports NVIDIA DGX SuperPOD and BasePOD deployments, Red Hat Enterprise Linux, Rocky, and Ubuntu operating systems.\u003c\/p\u003e\n\u003ch3 style=\"margin:1.25em 0 0.5em\"\u003eWarranty\u003c\/h3\u003e\n\u003cp style=\"margin:1em 0\"\u003eCovered under a 3-YEAR EXTENDED WARRANTY from the manufacturer.\u003c\/p\u003e\n\u003ch3 style=\"margin:1.25em 0 0.5em\"\u003eWhat's included\u003c\/h3\u003e\n\u003cul style=\"margin:1em 0\"\u003e\n\u003cli\u003eNVIDIA DGX B300 server\u003c\/li\u003e\n\u003cli\u003ePower cables\u003c\/li\u003e\n\u003cli\u003eDocumentation\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ctable class=\"pu-table\" style=\"width:100%;border-collapse:separate;border-spacing:0;border:0;box-shadow:none;outline:none;background:none;table-layout:fixed;margin:0\"\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd style=\"border:0;padding:0\"\u003e\n\u003ch3 style=\"margin:1.25em 0 0.5em\"\u003eFrequently asked questions\u003c\/h3\u003e\n\u003cstyle\u003e.pu-faq[open]\u003esummary .pu-faq-c{transform:rotate(180deg)}\u003c\/style\u003e\n\u003ctable class=\"pu-table\" style=\"width:100%;border-collapse:separate;border-spacing:0;border:0;box-shadow:none;outline:none;background:none;table-layout:fixed;margin:0 0 1em\"\u003e\u003ctbody\u003e\n\u003ctr\u003e\u003ctd style=\"border:0;padding:0;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e\u003cdetails class=\"pu-faq\" style=\"padding:0.45em 0\"\u003e\u003csummary style=\"cursor:pointer;display:flex;align-items:center;justify-content:space-between;gap:10px;list-style:none\"\u003e\u003cstrong\u003eWhat software comes pre-installed on the NVIDIA DGX B300?\u003c\/strong\u003e\u003csvg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewbox=\"0 0 24 24\" width=\"18\" height=\"18\" fill=\"none\" stroke=\"currentColor\" stroke-width=\"1.75\" stroke-linecap=\"round\" stroke-linejoin=\"round\" class=\"pu-faq-c\" style=\"width:1em;height:1em;flex:none;opacity:0.6;transition:transform 0.15s\" aria-hidden=\"true\"\u003e\u003cpath d=\"M6 9l6 6l6 -6\"\u003e\u003c\/path\u003e\u003c\/svg\u003e\u003c\/summary\u003e\u003cp style=\"margin:0.4em 0 0\"\u003eIt ships with NVIDIA AI Enterprise, NVIDIA Mission Control, and NVIDIA DGX OS, and supports Red Hat Enterprise Linux, Rocky, and Ubuntu.\u003c\/p\u003e\u003c\/details\u003e\u003c\/td\u003e\u003c\/tr\u003e\n\u003ctr\u003e\u003ctd style=\"border:0;padding:0;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e\u003cdetails class=\"pu-faq\" style=\"padding:0.45em 0\"\u003e\u003csummary style=\"cursor:pointer;display:flex;align-items:center;justify-content:space-between;gap:10px;list-style:none\"\u003e\u003cstrong\u003eDoes the NVIDIA DGX B300 support liquid cooling?\u003c\/strong\u003e\u003csvg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewbox=\"0 0 24 24\" width=\"18\" height=\"18\" fill=\"none\" stroke=\"currentColor\" stroke-width=\"1.75\" stroke-linecap=\"round\" stroke-linejoin=\"round\" class=\"pu-faq-c\" style=\"width:1em;height:1em;flex:none;opacity:0.6;transition:transform 0.15s\" aria-hidden=\"true\"\u003e\u003cpath d=\"M6 9l6 6l6 -6\"\u003e\u003c\/path\u003e\u003c\/svg\u003e\u003c\/summary\u003e\u003cp style=\"margin:0.4em 0 0\"\u003eYes, the platform is available in both air-cooled (8U) and liquid-cooled (4U) configurations, with the liquid-cooled variant using direct-to-chip technology.\u003c\/p\u003e\u003c\/details\u003e\u003c\/td\u003e\u003c\/tr\u003e\n\u003ctr\u003e\u003ctd style=\"border:0;padding:0;border-bottom:1px solid rgba(128,128,128,0.16);border-bottom:1px solid color-mix(in srgb,currentColor 12%,transparent)\"\u003e\u003cdetails class=\"pu-faq\" style=\"padding:0.45em 0\"\u003e\u003csummary style=\"cursor:pointer;display:flex;align-items:center;justify-content:space-between;gap:10px;list-style:none\"\u003e\u003cstrong\u003eWhat management network interfaces are included?\u003c\/strong\u003e\u003csvg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewbox=\"0 0 24 24\" width=\"18\" height=\"18\" fill=\"none\" stroke=\"currentColor\" stroke-width=\"1.75\" stroke-linecap=\"round\" stroke-linejoin=\"round\" class=\"pu-faq-c\" style=\"width:1em;height:1em;flex:none;opacity:0.6;transition:transform 0.15s\" aria-hidden=\"true\"\u003e\u003cpath d=\"M6 9l6 6l6 -6\"\u003e\u003c\/path\u003e\u003c\/svg\u003e\u003c\/summary\u003e\u003cp style=\"margin:0.4em 0 0\"\u003eIt includes a 1GbE onboard network interface card with RJ45 and a 1GbE RJ45 host baseboard management controller.\u003c\/p\u003e\u003c\/details\u003e\u003c\/td\u003e\u003c\/tr\u003e\n\u003ctr\u003e\u003ctd style=\"border:0;padding:0\"\u003e\u003cdetails class=\"pu-faq\" style=\"padding:0.45em 0\"\u003e\u003csummary style=\"cursor:pointer;display:flex;align-items:center;justify-content:space-between;gap:10px;list-style:none\"\u003e\u003cstrong\u003eWhat storage configuration does the NVIDIA DGX B300 use?\u003c\/strong\u003e\u003csvg xmlns=\"http:\/\/www.w3.org\/2000\/svg\" viewbox=\"0 0 24 24\" width=\"18\" height=\"18\" fill=\"none\" stroke=\"currentColor\" stroke-width=\"1.75\" stroke-linecap=\"round\" stroke-linejoin=\"round\" class=\"pu-faq-c\" style=\"width:1em;height:1em;flex:none;opacity:0.6;transition:transform 0.15s\" aria-hidden=\"true\"\u003e\u003cpath d=\"M6 9l6 6l6 -6\"\u003e\u003c\/path\u003e\u003c\/svg\u003e\u003c\/summary\u003e\u003cp style=\"margin:0.4em 0 0\"\u003eIt includes two 1.9 TB NVMe M.2 drives for the OS and eight 3.84 TB NVMe E1.S drives for internal storage.\u003c\/p\u003e\u003c\/details\u003e\u003c\/td\u003e\u003c\/tr\u003e\n\u003c\/tbody\u003e\u003c\/table\u003e\n\u003c\/td\u003e\u003c\/tr\u003e\u003c\/tbody\u003e\u003c\/table\u003e","brand":"NVIDIA","offers":[{"title":"Default Title","offer_id":53547829362999,"sku":"D0B3-G2304+P2CMI36","price":277200.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0999\/0841\/9895\/files\/EXX-IMG-9371902.webp?v=1793975826","url":"https:\/\/ciou.shop\/products\/nvidia-dgx-b300-ai-server-8x-blackwell-ultra-gpu","provider":"Ciou","version":"1.0","type":"link"}