{"product_id":"nvidia-rtx-pro-5000-blackwell-72gb","title":"NVIDIA RTX PRO 5000 Blackwell 72GB","description":"\u003csection class=\"phd-intro\"\u003e\n  \u003cdiv\u003e\n\u003cp class=\"phd-eyebrow\"\u003eWorkstation graphics \/ NVIDIA Blackwell\u003c\/p\u003e\n\u003ch2\u003e72GB for memory-intensive workstation workloads.\u003c\/h2\u003e\n\u003c\/div\u003e\n  \u003cdiv\u003e\n\u003cp\u003eThe NVIDIA RTX PRO 5000 72GB Blackwell adds 24GB of memory over the 48GB version for local AI and GPU data workflows that need more room. Its 72GB of ECC-capable GDDR7 retains the same 1,344GB\/s bandwidth and 300W, actively cooled dual-slot format.\u003c\/p\u003e\n\u003cp\u003eA full-height, dual-slot graphics card with active cooling and four DisplayPort 2.1b outputs.\u003c\/p\u003e\n\u003c\/div\u003e\n\u003c\/section\u003e\n\u003cnav class=\"phd-jump-links\" aria-label=\"Product information\"\u003e\n  \u003ca href=\"#hardware-components\"\u003eProduct details\u003c\/a\u003e\n  \u003ca href=\"#hardware-specification\"\u003eTechnical specification\u003c\/a\u003e\n  \u003ca href=\"#hardware-included\"\u003eIncluded and required\u003c\/a\u003e\n  \u003ca href=\"#hardware-questions\"\u003eBefore you order\u003c\/a\u003e\n\u003c\/nav\u003e\n\u003cdiv class=\"phd-metrics\" aria-label=\"Key specifications\"\u003e\n  \u003cdiv\u003e\n\u003cstrong\u003e72GB\u003c\/strong\u003e\u003cspan\u003eGDDR7 memory with ECC\u003c\/span\u003e\n\u003c\/div\u003e\n  \u003cdiv\u003e\n\u003cstrong\u003e1,344GB\/s\u003c\/strong\u003e\u003cspan\u003eGPU memory bandwidth\u003c\/span\u003e\n\u003c\/div\u003e\n  \u003cdiv\u003e\n\u003cstrong\u003e300W\u003c\/strong\u003e\u003cspan\u003eTotal board power\u003c\/span\u003e\n\u003c\/div\u003e\n  \u003cdiv\u003e\n\u003cstrong\u003e4x DisplayPort\u003c\/strong\u003e\u003cspan\u003e2.1b display outputs\u003c\/span\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003csection class=\"phd-components\" id=\"hardware-components\"\u003e\n  \u003ch2\u003eMore memory for demanding projects.\u003c\/h2\u003e\n  \u003cdiv class=\"phd-features\"\u003e\n    \u003carticle\u003e\u003ch3\u003eMore room for local AI and data\u003c\/h3\u003e\n\u003cp\u003e72GB provides additional space for model weights, LLM context and GPU-resident datasets rather than a general speed increase over the 48GB card. Fifth-generation Tensor Cores support FP4 inference in compatible software, while CUDA-X libraries support GPU data preparation and exploration. Allow for working buffers as well as the data itself.\u003c\/p\u003e\u003c\/article\u003e\n    \u003carticle\u003e\u003ch3\u003eCapacity for detailed 3D scenes\u003c\/h3\u003e\n\u003cp\u003eThe extra memory can accommodate more scene geometry and textures when a rendering project is constrained by VRAM. Fourth-generation RT cores accelerate ray-traced lighting and reflections for CAD visualisation and design reviews; supported neural-rendering features can assist interactive previews.\u003c\/p\u003e\u003c\/article\u003e\n    \u003carticle\u003e\u003ch3\u003eEdit, transcode and deliver video\u003c\/h3\u003e\n\u003cp\u003eThree ninth-generation NVENC and three sixth-generation NVDEC engines handle video encoding and decoding for editing, transcoding and streaming. H.264 and HEVC 4:2:2 support suits compatible camera footage, with AV1 encoding available for delivery in supported software. Four DisplayPort 2.1b outputs support a multi-monitor editing workspace.\u003c\/p\u003e\u003c\/article\u003e\n  \u003c\/div\u003e\n\u003c\/section\u003e\n\u003caside class=\"phd-note\"\u003e\u003ch3\u003eCheck space, power and cooling.\u003c\/h3\u003e\n\u003cp\u003eAllow for the full-height, dual-slot card, its 10.5-inch length and the PCIe CEM5 16-pin power cable. Check that the chassis provides suitable airflow and that the PSU can support the 300W card alongside the rest of the system.\u003c\/p\u003e\u003c\/aside\u003e\n\u003csection class=\"phd-specs\" id=\"hardware-specification\"\u003e\n  \u003cdiv class=\"phd-section-heading\"\u003e\n\u003ch2\u003eTechnical specification\u003c\/h2\u003e\n\u003cp\u003eSpecifications for one card\u003c\/p\u003e\n\u003c\/div\u003e\n  \u003ctable\u003e\n\u003ccaption\u003e01 \/ Architecture, memory and compute\u003c\/caption\u003e\n\u003ctbody\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eGPU model\u003c\/th\u003e\n\u003ctd\u003eNVIDIA RTX PRO 5000 72GB Blackwell\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eNVIDIA part number\u003c\/th\u003e\n\u003ctd\u003e900-5G153-2270-000\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eArchitecture\u003c\/th\u003e\n\u003ctd\u003eNVIDIA Blackwell\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eGPU memory\u003c\/th\u003e\n\u003ctd\u003e72GB GDDR7 with ECC\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eMemory bandwidth\u003c\/th\u003e\n\u003ctd\u003e1,344GB\/s\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eCUDA cores\u003c\/th\u003e\n\u003ctd\u003e14,080\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eTensor Cores\u003c\/th\u003e\n\u003ctd\u003e440, fifth generation\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eRT cores\u003c\/th\u003e\n\u003ctd\u003e110, fourth generation\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eMulti-Instance GPU\u003c\/th\u003e\n\u003ctd\u003eUp to 2x 36GB, or 1x 72GB; software and firmware prerequisites apply\u003c\/td\u003e\n\u003c\/tr\u003e\n  \u003ctr\u003e\n\u003cth scope=\"row\"\u003eFP32 single-precision peak\u003c\/th\u003e\n\u003ctd\u003e65 TFLOPS\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003cth scope=\"row\"\u003eFP4 Tensor peak (with sparsity)\u003c\/th\u003e\n\u003ctd\u003e2,064 TOPS\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\n\u003cp class=\"phd-footnote\"\u003eManufacturer peaks per GPU at GPU Boost clock, not application benchmarks. FP4 uses sparsity; it is not dense throughput. Actual performance depends on the application, model and software support.\u003c\/p\u003e\n  \u003ctable\u003e\n\u003ccaption\u003e02 \/ Connectivity and video\u003c\/caption\u003e\n\u003ctbody\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eHost interface\u003c\/th\u003e\n\u003ctd\u003ePCIe 5.0 x16\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003ePhysical display outputs\u003c\/th\u003e\n\u003ctd\u003e4x DisplayPort 2.1b\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eDirectly connected displays\u003c\/th\u003e\n\u003ctd\u003eUp to 4; supported mode depends on the display configuration\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eVideo engines\u003c\/th\u003e\n\u003ctd\u003e3x ninth-generation NVENC; 3x sixth-generation NVDEC\u003c\/td\u003e\n\u003c\/tr\u003e\n  \u003c\/tbody\u003e\n\u003c\/table\u003e\n  \u003ctable\u003e\n\u003ccaption\u003e03 \/ Power, cooling and physical format\u003c\/caption\u003e\n\u003ctbody\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eBoard power\u003c\/th\u003e\n\u003ctd\u003e300W total board power\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eAuxiliary power input\u003c\/th\u003e\n\u003ctd\u003e1x PCIe CEM5 16-pin\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eCooling\u003c\/th\u003e\n\u003ctd\u003eActive blower fan\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eForm factor\u003c\/th\u003e\n\u003ctd\u003eFull height, dual slot\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eNominal card dimensions\u003c\/th\u003e\n\u003ctd\u003e4.4 inches high x 10.5 inches long\u003c\/td\u003e\n\u003c\/tr\u003e\n  \u003c\/tbody\u003e\n\u003c\/table\u003e\n  \u003cp class=\"phd-footnote\"\u003eSize the PSU for the complete system, including the 300W graphics card. Allow additional clearance beyond the nominal card dimensions for power cabling and airflow.\u003c\/p\u003e\n\u003c\/section\u003e\n\u003csection class=\"phd-supply\" id=\"hardware-included\"\u003e\n  \u003cdiv\u003e\n\u003cp class=\"phd-eyebrow\"\u003eIncluded\u003c\/p\u003e\n\u003ch2\u003eOne 72GB graphics card.\u003c\/h2\u003e\n\u003cp\u003eOne NVIDIA RTX PRO 5000 72GB Blackwell graphics card, part number 900-5G153-2270-000.\u003c\/p\u003e\n\u003cp\u003ePlan any power and display cables, installation accessories and software licences separately.\u003c\/p\u003e\n\u003c\/div\u003e\n  \u003cdiv\u003e\n\u003cp class=\"phd-eyebrow\"\u003ePlan your installation\u003c\/p\u003e\n\u003ch2\u003eWhat your workstation needs.\u003c\/h2\u003e\n\u003cul\u003e\n\u003cli\u003eA compatible full-height PCIe x16 slot with dual-slot clearance.\u003c\/li\u003e\n\u003cli\u003eSystem power capacity appropriate to a 300W GPU and the remaining components, with approved PCIe CEM5 16-pin cabling.\u003c\/li\u003e\n\u003cli\u003eClearance for the card and power connection, with appropriate chassis airflow.\u003c\/li\u003e\n\u003cli\u003eA supported NVIDIA driver, operating system and application stack; additional licences and display cables where required.\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003c\/div\u003e\n\u003c\/section\u003e\n\u003csection class=\"phd-faq\" id=\"hardware-questions\" aria-label=\"Before you order\"\u003e\n  \u003cdetails\u003e\u003csummary\u003eWhen is the extra memory useful?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eThe 72GB model provides 24GB more memory than the 48GB version, giving additional capacity for models, scenes and datasets that need more GPU memory. Both models have the same core counts, 1,344GB\/s memory bandwidth and 300W board power. Choose between them according to your application's memory requirements rather than expecting a general speed increase.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eWhat should I check for local AI workloads?\u003c\/summary\u003e\u003cdiv\u003e\n\u003cp\u003eCheck that your software supports the GPU and estimate its total memory requirement. Model precision, context length, concurrent requests and working buffers all affect VRAM use; the model file size alone is not enough to size the workload.\u003c\/p\u003e\n\u003cp\u003eMIG requires a supported Linux driver and vBIOS configuration. Check display-mode requirements before planning connected monitors.\u003c\/p\u003e\n\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eWhat does my workstation need?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eA compatible full-height PCIe x16 slot, two-slot clearance and room for the 10.5-inch card and power cable. The card uses one PCIe CEM5 16-pin auxiliary power connection and has a 300W board rating. Check PSU capacity for the whole system and provide suitable chassis airflow.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eCan I connect four monitors?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eYes. The card has four DisplayPort 2.1b outputs and supports up to four simultaneous displays. Check the resolution, refresh rate and colour format you need against the capabilities of the displays, cables and software.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eWhat is included?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eOne RTX PRO 5000 72GB Blackwell graphics card. Plan any power adapters, display cables, installation accessories and software licences separately.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\n  \n  \n  \n  \n  \u003cdetails\u003e\u003csummary\u003eManufacturer documentation\u003c\/summary\u003e\u003cdiv\u003e\n\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.nvidia.com\/content\/dam\/en-zz\/Solutions\/products\/workstations\/professional-desktop-gpus\/rtx-pro-5000-blackwell\/workstation-datasheet-blackwell-rtx-pro-5000-gtc25-spring-nvidia-3658700.pdf\"\u003eNVIDIA RTX PRO 5000 Blackwell paired 48GB\/72GB datasheet\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.pny.com\/File%20Library\/Company\/Support\/Product%20Brochures\/NVIDIA%20RTX\/English\/nvidia-rtx-pro-5000-blackwell.pdf\"\u003ePNY RTX PRO 5000 Blackwell paired datasheet\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.pny.com\/nvidia-rtx-pro-5000-72gb-blackwell\"\u003ePNY RTX PRO 5000 72GB Blackwell specifications\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.pny.com\/File%20Library\/Company\/Support\/linecards\/pro-viz-gpus\/nvidia-professional-graphics-linecard.pdf\"\u003ePNY professional graphics linecard - exact 72GB part number\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/docs.nvidia.com\/datacenter\/tesla\/mig-user-guide\/deployment-considerations.html\"\u003eNVIDIA MIG deployment considerations\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/docs.nvidia.com\/datacenter\/tesla\/mig-user-guide\/getting-started-with-mig.html\"\u003eNVIDIA MIG setup and firmware prerequisites\u003c\/a\u003e\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003c\/div\u003e\u003c\/details\u003e\n\u003c\/section\u003e\n","brand":"NVIDIA","offers":[{"title":"Default Title","offer_id":60872452014414,"sku":"900-5G153-2270-000","price":9495.0,"currency_code":"GBP","in_stock":false}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1078\/0213\/2814\/files\/5000-blackwell-official-pny-front.png?v=1789853855","url":"https:\/\/shop.perceptasolutions.com\/products\/nvidia-rtx-pro-5000-blackwell-72gb","provider":"Percepta Solutions","version":"1.0","type":"link"}