{"product_id":"minisforum-ai-x1-pro-barebone","title":"MINISFORUM AI X1 Pro Barebone Mini PC Ryzen AI 9 HX 370 | No RAM\/SSD","description":"\u003cdiv class=\"mg-desc\"\u003e\n\u003cp\u003eThe \u003cstrong\u003eMINISFORUM AI X1 Pro barebone\u003c\/strong\u003e is a \u003cstrong\u003emini PC chassis with the AMD Ryzen AI 9 HX 370 processor\u003c\/strong\u003e (12 cores, 24 threads, up to 5.1 GHz), Radeon 890M graphics, and a 50 TOPS NPU, shipped without memory, storage, or an operating system. You add two DDR5-5600 SO-DIMMs (up to 128 GB total) and up to three M.2 2280 NVMe SSDs. It has dual 2.5 Gigabit Ethernet ports, WiFi 7, Bluetooth 5.4, two USB4 ports, and an OCuLink port for an external GPU. It is for people who already own RAM and an SSD, who want to pick their own parts, or who plan to run Linux and a local AI stack on hardware they configured themselves.\u003c\/p\u003e\n\n\u003ch2\u003eWho the AI X1 Pro barebone is for\u003c\/h2\u003e\n\u003cul\u003e\n\u003cli\u003eLocal AI builders who want to fill both SO-DIMM slots with 64 GB or 96 GB and run larger quantized models through Ollama or similar tools\u003c\/li\u003e\n\u003cli\u003eLinux and home-server users who do not want to pay for a Windows license they will wipe on day one\u003c\/li\u003e\n\u003cli\u003eAnyone with spare DDR5 SO-DIMMs and NVMe drives from a laptop upgrade\u003c\/li\u003e\n\u003cli\u003eSmall offices that want a quiet desktop with dual 2.5 GbE and a fingerprint reader, configured to a specific memory and storage spec\u003c\/li\u003e\n\u003cli\u003eGamers and creators who plan to attach an external GPU over OCuLink later\u003c\/li\u003e\n\u003c\/ul\u003e\n\n\u003ch2\u003eKey features\u003c\/h2\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eAMD Ryzen AI 9 HX 370\u003c\/strong\u003e: 12 cores, 24 threads, 2.0 GHz base, up to 5.1 GHz boost\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eRadeon 890M integrated graphics\u003c\/strong\u003e at 2900 MHz, sharing system memory with the CPU\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e50 TOPS NPU\u003c\/strong\u003e (80 TOPS combined across CPU, GPU, and NPU per MINISFORUM)\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eTwo DDR5 SO-DIMM slots\u003c\/strong\u003e rated 5600 MT\/s, up to 128 GB total (sold empty)\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eThree M.2 2280 NVMe slots\u003c\/strong\u003e: two at PCIe 4.0 x4 and one at PCIe 4.0 x1, up to 4 TB per slot (sold empty)\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eQuad display output\u003c\/strong\u003e: HDMI 2.1 (8K at 60 Hz or 4K at 120 Hz), DisplayPort 2.0 (8K at 60 Hz or 4K at 120 Hz), and two USB4 ports (4K at 60 Hz each)\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eDual 2.5 GbE RJ-45 ports\u003c\/strong\u003e for a router, NAS, or a separate management network\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eWiFi 7 and Bluetooth 5.4\u003c\/strong\u003e built in\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eOCuLink port\u003c\/strong\u003e (PCIe 4.0 x4) for an external graphics card dock\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eBuilt-in power supply\u003c\/strong\u003e (19 V, 7.1 A, 134.9 W), so there is no external power brick\u003c\/li\u003e\n\u003cli\u003e\u003cstrong\u003eFingerprint reader, SD card slot, 3.5 mm combo jack, dual speakers, dual microphones, Kensington lock\u003c\/strong\u003e\u003c\/li\u003e\n\u003c\/ul\u003e\n\n\u003ch2\u003eSpecifications\u003c\/h2\u003e\n\u003ctable class=\"mg-spec-table\"\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eModel\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003eMINISFORUM AI X1 Pro, barebone (no RAM, no SSD, no OS)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eProcessor\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003eAMD Ryzen AI 9 HX 370, 12C\/24T, 2.0 GHz base, 5.1 GHz max\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eGraphics\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003eAMD Radeon 890M, 2900 MHz\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eNPU\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e50 TOPS NPU; 80 TOPS total platform\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eMemory\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e2 x DDR5 SO-DIMM, 5600 MT\/s, max 128 GB (empty)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eStorage\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e3 x M.2 2280 NVMe (2 x PCIe 4.0 x4, 1 x PCIe 4.0 x1), up to 4 TB each (empty)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eVideo out\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e1 x HDMI 2.1, 1 x DP 2.0, 2 x USB4; four 4K displays at once\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eUSB\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e2 x USB4 (40 Gbps; rear port PD-in 100 W), 2 x USB-A 3.2 Gen 2, 1 x USB-A 2.0\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eNetworking\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e2 x 2.5 GbE RJ-45; WiFi 7; Bluetooth 5.4\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eExpansion\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e1 x OCuLink (PCIe 4.0 x4); 1 x SD card slot\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003ePower\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003eInternal supply, 19 V \/ 7.1 A (134.9 W), AC inlet\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eSpec\u003c\/span\u003eSize and weight\u003c\/td\u003e\n\u003ctd\u003e\n\u003cspan class=\"mg-td-label\"\u003eDetail\u003c\/span\u003e195 x 195 x 42.5 mm (47.5 mm with feet); 1.5 kg\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\n\n\u003ch2\u003eWhat \"barebone\" means\u003c\/h2\u003e\n\u003cp\u003eA barebone PC is a finished case, motherboard, processor, cooler, and power supply with no memory, no drive, and no operating system installed. You open the chassis, seat two DDR5 SO-DIMM modules and at least one M.2 2280 NVMe SSD, then install Windows or Linux from a USB stick. MINISFORUM sells the same AI X1 Pro preconfigured; the barebone is the route for buyers who want to choose exact part brands, pair 64 GB or 96 GB kits for local AI, or avoid paying for an OS they will replace.\u003c\/p\u003e\n\n\u003cdiv class=\"mg-cross\"\u003e\n\u003cp class=\"mg-cross-kicker\"\u003ePrefer it ready to run?\u003c\/p\u003e\n\u003cp class=\"mg-cross-text\"\u003eThe \u003ca href=\"\/products\/minisforum-ai-x1-pro-370\"\u003eAI X1 Pro-370 configured model\u003c\/a\u003e is the same chassis with 32 GB of DDR5 and a 1 TB PCIe 4.0 SSD installed. Pick it if you want to plug in and boot; pick the barebone if you want to choose your own memory and storage.\u003c\/p\u003e\n\u003c\/div\u003e\n\n\u003ch2\u003eRunning local AI on the AI X1 Pro\u003c\/h2\u003e\n\u003cp\u003eThe Radeon 890M has no dedicated video memory; it draws from the same DDR5 pool as the CPU, so the amount of RAM you install sets the largest model you can load. A rough rule: a 4-bit quantized model needs a little over half its parameter count in gigabytes (a 14B model needs roughly 9 GB, a 32B model roughly 20 GB, a 70B model roughly 40 GB) plus a few gigabytes for context and the operating system. With 32 GB installed, 7B to 14B models run with room to spare and 32B models fit at 4-bit. With 64 GB or more, 70B models at 4-bit fit in memory, though they run far slower than on a discrete GPU. Our \u003ca href=\"\/blogs\/modemguides-blog\/best-mini-pc-local-ai-ollama-2026\"\u003emini PC tier guide for local AI\u003c\/a\u003e places this class of hardware in context, and \u003ca href=\"\/blogs\/modemguides-blog\/can-a-mini-pc-run-a-70b-model\"\u003ethis guide on whether a mini PC can run a 70B model\u003c\/a\u003e covers memory math and realistic expectations. The OCuLink port is the upgrade path if you later want a discrete GPU for faster inference.\u003c\/p\u003e\n\n\u003ch2\u003eCommon questions\u003c\/h2\u003e\n\u003ch3\u003eHow much RAM can the AI X1 Pro take?\u003c\/h3\u003e\n\u003cp\u003eUp to 128 GB across two DDR5 SO-DIMM slots at 5600 MT\/s, per MINISFORUM's specification. Use a matched pair for dual-channel bandwidth, which matters for both the Radeon 890M and local AI inference.\u003c\/p\u003e\n\u003ch3\u003eCan the AI X1 Pro run local AI models?\u003c\/h3\u003e\n\u003cp\u003eYes. The Ryzen AI 9 HX 370 runs Ollama, LM Studio, and similar tools on the CPU and integrated GPU. The model size you can load depends on installed RAM: 32 GB covers up to roughly 32B models at 4-bit, and 64 GB or more covers 70B models at 4-bit. Larger models run slowly without a discrete GPU.\u003c\/p\u003e\n\u003ch3\u003eDoes the barebone come with Windows?\u003c\/h3\u003e\n\u003cp\u003eNo. MINISFORUM lists the barebone configuration as no RAM, no SSD, and no operating system. You install Windows 11 or a Linux distribution yourself.\u003c\/p\u003e\n\u003ch3\u003eHow many SSDs fit in the AI X1 Pro?\u003c\/h3\u003e\n\u003cp\u003eThree M.2 2280 NVMe drives: two slots at PCIe 4.0 x4 and one at PCIe 4.0 x1, up to 4 TB each.\u003c\/p\u003e\n\u003ch3\u003eDoes the AI X1 Pro have a 2.5 GbE port?\u003c\/h3\u003e\n\u003cp\u003eYes, two of them. Pair one with a \u003ca href=\"\/collections\/switches\"\u003e2.5 GbE switch\u003c\/a\u003e or a router with a multi-gig LAN port to move large model files and backups at full speed.\u003c\/p\u003e\n\u003ch3\u003eCan I add a graphics card?\u003c\/h3\u003e\n\u003cp\u003eYes, through the OCuLink port (PCIe 4.0 x4) with an external GPU dock, or over USB4 with a Thunderbolt-style enclosure. OCuLink is a cable standard that carries PCIe lanes outside the case.\u003c\/p\u003e\n\u003c\/div\u003e\n\n\u003cstyle\u003e\n.mg-desc p,.mg-desc li{font-size:17px !important;line-height:1.6 !important;color:inherit;}\n.mg-desc li{margin-bottom:6px;}\n.mg-desc h2{font-size:24px !important;margin:28px 0 10px !important;color:inherit;}\n.mg-desc h3{font-size:19px !important;margin:20px 0 6px !important;color:inherit;}\n.mg-desc .mg-spec-table{width:100%;border-collapse:collapse;margin:14px 0 22px;font-size:16px;color:inherit;}\n.mg-desc .mg-spec-table th,.mg-desc .mg-spec-table td{padding:10px 12px;border:1px solid rgba(128,128,128,0.3);text-align:left;vertical-align:top;line-height:1.5;color:inherit;}\n.mg-desc .mg-spec-table th{background:rgba(10,94,166,0.08);font-weight:700;}\n.mg-desc .mg-spec-table tbody tr:nth-child(even){background:rgba(128,128,128,0.06);}\n.mg-desc .mg-spec-table td:first-child{font-weight:600;width:34%;}\n.mg-desc .mg-td-label{display:none;}\n.mg-desc .mg-cross{border:2px solid rgba(179,164,134,0.6);border-left:6px solid #b3a486;border-radius:10px;padding:14px 18px;margin:24px 0;background:rgba(179,164,134,0.06);color:inherit;}\n.mg-desc .mg-cross .mg-cross-kicker{font-size:13px !important;letter-spacing:0.08em;text-transform:uppercase;opacity:0.75;margin:0 0 6px !important;font-weight:700;color:inherit;}\n.mg-desc .mg-cross .mg-cross-text{margin:0 !important;font-size:17px !important;line-height:1.5 !important;color:inherit;}\n.mg-desc .mg-cross a{font-weight:700;color:inherit;text-decoration:underline;}\n@media (max-width:768px){\n  .mg-desc p,.mg-desc li{font-size:16px !important;}\n  .mg-desc h2{font-size:22px !important;}\n  .mg-desc .mg-spec-table{font-size:15px;}\n}\n\u003c\/style\u003e","brand":"MINISFORUM","offers":[{"title":"Default Title","offer_id":50677377302614,"sku":null,"price":695.0,"currency_code":"USD","in_stock":false}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0582\/9463\/2534\/files\/CopyofUntitledDesign_17.png?v=1793377285","url":"https:\/\/www.modemguides.com\/products\/minisforum-ai-x1-pro-barebone","provider":"ModemGuides","version":"1.0","type":"link"}