{"product_id":"amd-radeon-ai-pro-r9700-32gb-asus-turbo","title":"AMD Radeon AI PRO R9700 32GB (ASUS Turbo)","description":"\u003csection class=\"phd-intro\"\u003e\n  \u003cdiv\u003e\n\u003cp class=\"phd-eyebrow\"\u003eWorkstation graphics \/ AMD RDNA 4\u003c\/p\u003e\n\u003ch2\u003e32GB of GDDR6 for local AI, in a blower-cooled card.\u003c\/h2\u003e\n\u003c\/div\u003e\n  \u003cdiv\u003e\n\u003cp\u003eThe ASUS Turbo Radeon AI PRO R9700 puts AMD's RDNA 4 workstation GPU and 32GB of memory on a full-height, dual-slot card. It suits local LLM inference, AI development with AMD ROCm and professional graphics. Its blower cooler pushes most of its heat out through the rear bracket, a design AMD describes as suited to workstations running several cards side by side.\u003c\/p\u003e\n\u003cp\u003eThis is the ASUS board, part number 90YV0MN0-M0NA00.\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\u003e32GB\u003c\/strong\u003e\u003cspan\u003eGDDR6 memory\u003c\/span\u003e\n\u003c\/div\u003e\n  \u003cdiv\u003e\n\u003cstrong\u003e640GB\/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\u003eUp to 4\u003c\/strong\u003e\u003cspan\u003eCards with ROCm multi-GPU on Linux\u003c\/span\u003e\n\u003c\/div\u003e\n\u003c\/div\u003e\n\u003csection class=\"phd-components\" id=\"hardware-components\"\u003e\n  \u003ch2\u003eMemory, AI accelerators and room for more cards.\u003c\/h2\u003e\n  \u003cdiv class=\"phd-features\"\u003e\n    \u003carticle\u003e\u003ch3\u003eLocal AI with ROCm\u003c\/h3\u003e\n\u003cp\u003eOn Linux, AMD ROCm supports the R9700 with PyTorch, TensorFlow and vLLM, plus JAX and llama.cpp for inference. On Windows 11, PyTorch runs for inference, and desktop LLM apps such as LM Studio run through Vulkan. The 32GB memory budget has to cover the model, its context and working buffers.\u003c\/p\u003e\u003c\/article\u003e\n    \u003carticle\u003e\u003ch3\u003eSecond-generation AI accelerators\u003c\/h3\u003e\n\u003cp\u003e128 AI accelerators handle FP16, FP8 and INT8 matrix work. AMD rates the card at 191 TFLOPS FP16 and 383 TOPS INT8 for matrix operations, doubling with structured sparsity. PyTorch on ROCm uses the familiar torch.cuda interface, so many PyTorch scripts run without code changes.\u003c\/p\u003e\u003c\/article\u003e\n    \u003carticle\u003e\u003ch3\u003eBuilt for multi-card workstations\u003c\/h3\u003e\n\u003cp\u003eAMD's ROCm multi-GPU support on Linux covers up to four cards. There is no bridge: the cards communicate over PCIe, and AMD requires CPU-attached slots with identical lane widths that support PCIe atomics. The blower cooler exhausts most of its heat out through the rear bracket, so cards can sit next to each other in a case with good airflow.\u003c\/p\u003e\u003c\/article\u003e\n  \u003c\/div\u003e\n\u003c\/section\u003e\n\u003caside class=\"phd-note\"\u003e\u003ch3\u003eCheck the 16-pin power cable and PSU.\u003c\/h3\u003e\n\u003cp\u003eThe card takes one 16-pin 12V-2x6 power connector at the far end of the card, away from the bracket. AMD recommends at least a 750W PSU for one card. Allow room for the 266.7mm card and the power cable behind it.\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 \/ GPU and memory\u003c\/caption\u003e\n\u003ctbody\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eGPU\u003c\/th\u003e\n\u003ctd\u003eAMD Radeon AI PRO R9700, RDNA 4 architecture\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eBoard\u003c\/th\u003e\n\u003ctd\u003eASUS Turbo, part number 90YV0MN0-M0NA00\u003c\/td\u003e\n\u003c\/tr\u003e\n        \u003ctr\u003e\n\u003cth scope=\"row\"\u003eCompute units\u003c\/th\u003e\n\u003ctd\u003e64\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eStream processors\u003c\/th\u003e\n\u003ctd\u003e4,096\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eAI accelerators\u003c\/th\u003e\n\u003ctd\u003e128, second generation\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eRay accelerators\u003c\/th\u003e\n\u003ctd\u003e64\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eGPU memory\u003c\/th\u003e\n\u003ctd\u003e32GB GDDR6; ECC supported on Linux\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eMemory interface\u003c\/th\u003e\n\u003ctd\u003e256-bit\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eMemory bandwidth\u003c\/th\u003e\n\u003ctd\u003e640GB\/s\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eAMD Infinity Cache\u003c\/th\u003e\n\u003ctd\u003e64MB\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eBoost clock\u003c\/th\u003e\n\u003ctd\u003eUp to 2,920MHz\u003c\/td\u003e\n\u003c\/tr\u003e\n  \u003c\/tbody\u003e\n\u003c\/table\u003e\n  \u003ctable\u003e\n\u003ccaption\u003e02 \/ AI and compute performance\u003c\/caption\u003e\n\u003ctbody\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eFP32 vector\u003c\/th\u003e\n\u003ctd\u003e47.8 TFLOPS\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eFP16 matrix\u003c\/th\u003e\n\u003ctd\u003e191 TFLOPS; 383 TFLOPS with structured sparsity\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eFP8 matrix\u003c\/th\u003e\n\u003ctd\u003e383 TFLOPS; 766 TFLOPS with structured sparsity\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eINT8 matrix\u003c\/th\u003e\n\u003ctd\u003e383 TOPS; 766 TOPS with structured sparsity\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eINT4 matrix\u003c\/th\u003e\n\u003ctd\u003e766 TOPS; 1,531 TOPS with structured sparsity\u003c\/td\u003e\n\u003c\/tr\u003e\n  \u003c\/tbody\u003e\n\u003c\/table\u003e\n  \u003cp class=\"phd-footnote\"\u003eAMD's peak figures for one GPU, not application benchmarks. Actual performance depends on the model, precision and software.\u003c\/p\u003e\n  \u003ctable\u003e\n\u003ccaption\u003e03 \/ Connectivity, video and software\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\"\u003eDisplay outputs\u003c\/th\u003e\n\u003ctd\u003e3x DisplayPort 2.1a, 1x HDMI 2.1b\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eMaximum displays\u003c\/th\u003e\n\u003ctd\u003e4\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eVideo engines\u003c\/th\u003e\n\u003ctd\u003eH.264, HEVC and AV1 encode and decode\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eMulti-GPU\u003c\/th\u003e\n\u003ctd\u003eUp to 4 cards with ROCm on Linux, over PCIe; no bridge\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eOperating systems\u003c\/th\u003e\n\u003ctd\u003eWindows 10 and 11 64-bit; Linux x86 64-bit\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eDrivers and software\u003c\/th\u003e\n\u003ctd\u003eAMD Software: PRO Edition (Windows); Radeon Software for Linux; AMD ROCm\u003c\/td\u003e\n\u003c\/tr\u003e\n  \u003c\/tbody\u003e\n\u003c\/table\u003e\n  \u003ctable\u003e\n\u003ccaption\u003e04 \/ Power, cooling and physical format\u003c\/caption\u003e\n\u003ctbody\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eTotal board power\u003c\/th\u003e\n\u003ctd\u003e300W (AMD specification)\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003ePower connector\u003c\/th\u003e\n\u003ctd\u003e1x 16-pin 12V-2x6, at the far end of the card\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eRecommended PSU\u003c\/th\u003e\n\u003ctd\u003e750W minimum for one card (AMD)\u003c\/td\u003e\n\u003c\/tr\u003e\n    \u003ctr\u003e\n\u003cth scope=\"row\"\u003eCooling\u003c\/th\u003e\n\u003ctd\u003eBlower fan, front-to-back airflow\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\"\u003eDimensions\u003c\/th\u003e\n\u003ctd\u003e266.7 x 111.1 x 40mm\u003c\/td\u003e\n\u003c\/tr\u003e\n  \u003c\/tbody\u003e\n\u003c\/table\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 graphics card.\u003c\/h2\u003e\n\u003cp\u003eOne ASUS Turbo Radeon AI PRO R9700 32GB graphics card. ASUS lists a quick setup guide in the box; no power adapter is listed.\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 PSU with a native 16-pin 12V-2x6 cable for each card: AMD recommends at least 750W for one card; for more cards, we suggest adding at least 300W per additional card.\u003c\/li\u003e\n\u003cli\u003eA full-height PCIe x16 slot with dual-slot clearance and room for a 266.7mm card and its power cable.\u003c\/li\u003e\n\u003cli\u003eFor several cards, CPU-attached slots with matching lane widths that support PCIe atomics.\u003c\/li\u003e\n\u003cli\u003eOn Windows, AMD Software: PRO Edition for professional applications, or the driver AMD specifies for PyTorch on Windows; on Linux, a supported distribution with AMD ROCm. Display cables as needed.\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\u003eWill it run my CUDA software?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eNot directly: the R9700 uses AMD ROCm rather than NVIDIA CUDA. PyTorch on ROCm keeps the torch.cuda interface, so many PyTorch projects run unchanged, and vLLM, TensorFlow and llama.cpp support ROCm on Linux. Check that your applications and any custom CUDA code support AMD GPUs.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eDoes it work on Windows?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eYes. AMD Software: PRO Edition supports the card for professional applications on Windows 10 and 11. For AI work on Windows 11, AMD supports PyTorch for inference, not training, using the graphics driver named in AMD's PyTorch-on-Windows guide, and LLM apps such as LM Studio can run through Vulkan. For training, the wider ROCm framework stack and multi-GPU work, use Linux.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eCan I run several cards together?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eYes. AMD's ROCm multi-GPU support on Linux covers up to four cards, and AMD advises running no more than two compute workloads at the same time; multi-GPU is not supported under WSL. The cards communicate over PCIe, so AMD requires CPU-attached slots with identical lane widths that support PCIe atomics. Size the PSU from AMD's 750W minimum for one card, adding at least 300W for each additional card, with one native 16-pin cable per card.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eCan I use it in a server?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eAMD designs the Radeon AI PRO R9700 for workstations and does not recommend it for data centre use. For servers, choose a passively cooled data centre GPU.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\u003cdetails\u003e\u003csummary\u003eWhat power connection does it need?\u003c\/summary\u003e\u003cdiv\u003e\u003cp\u003eOne 16-pin 12V-2x6 connector at the far end of the card. Use a PSU with a native 16-pin GPU cable; AMD recommends at least 750W for a system with one card.\u003c\/p\u003e\u003c\/div\u003e\u003c\/details\u003e\n  \u003cdetails\u003e\u003csummary\u003eManufacturer documentation\u003c\/summary\u003e\u003cdiv\u003e\n\u003cul\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.amd.com\/en\/products\/graphics\/workstations\/radeon-ai-pro\/ai-9000-series\/amd-radeon-ai-pro-r9700.html\"\u003eAMD Radeon AI PRO R9700 specifications\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.asus.com\/uk\/motherboards-components\/graphics-cards\/turbo\/turbo-ai-pro-r9700-32g\/techspec\/\"\u003eASUS Turbo Radeon AI PRO R9700 32GB technical specifications\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.amd.com\/en\/support\/downloads\/drivers.html\/graphics\/radeon-ai-pro\/radeon-ai-pro-r9000-series\/amd-radeon-ai-pro-r9700.html\"\u003eAMD drivers and software for the R9700\u003c\/a\u003e\u003c\/li\u003e\n\u003cli\u003e\u003ca href=\"https:\/\/www.amd.com\/en\/blogs\/2025\/amd-radeon-ai-pro-r9700-to-be-available-in-workstation.html\"\u003eAMD: Radeon AI PRO R9700 in workstations, ROCm and multi-GPU\u003c\/a\u003e\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003c\/div\u003e\u003c\/details\u003e\n\u003c\/section\u003e\n","brand":"AMD","offers":[{"title":"Default Title","offer_id":60995469607246,"sku":"90YV0MN0-M0NA00","price":2095.0,"currency_code":"GBP","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1078\/0213\/2814\/files\/ASUS_20TURBO_20Radeon_E2_84_A2_20AI_20PRO_20R9700-_20rear_20angled_20view.png?v=1790670323","url":"https:\/\/shop.perceptasolutions.com\/products\/amd-radeon-ai-pro-r9700-32gb-asus-turbo","provider":"Percepta Solutions","version":"1.0","type":"link"}