{"product_id":"ai-8850-llm-accelerator-m2-kit-8gb-version-ax8850","title":"AI-8850 LLM Accelerator M.2 Kit 8GB Version (AX8850)","description":"\u003cp\u003e\u003cvideo id=\"video\" playsinline=\"\" muted=\"\" controls=\"controls\" class=\"video-container\" width=\"100%\" style=\"border-radius: 8px; box-shadow: 0 4px 12px rgba(0,0,0,0.08);\"\u003e\n  \u003csource type=\"video\/mp4\" src=\"https:\/\/m5stack-doc.oss-cn-shenzhen.aliyuncs.com\/1215\/AI-002-8G-LLM-8850-Kit-EN.mp4\"\u003e\u003c\/video\u003e\u003c\/p\u003e\n\u003cstyle\u003e\n\/* ===== Scoped Description Style ===== *\/\n.product-description.rte {\n  font-family: -apple-system, BlinkMacSystemFont, \"Segoe UI\", Roboto, \"Helvetica Neue\", Arial, sans-serif;\n  color: #2c2c2c;\n  line-height: 1.65;\n  font-size: 16px;\n}\n\n\/* ===== Headings ===== *\/\n.product-description.rte h2 {\n  font-size: 1.8em;\n  color: #0D47A1;\n  border-left: 4px solid #1f6fe5;\n  padding-left: 12px;\n  margin-top: 32px;\n  margin-bottom: 16px;\n}\n\n\/* ===== H3 Titles Optimization ===== *\/\n.product-description.rte h3 {\n  font-size: 1.4em;\n  color: #0D47A1;\n  margin-top: 24px;\n  margin-bottom: 12px;\n  font-weight: 500;\n  padding-bottom: 6px;\n  border-bottom: 2px solid #e6e6e6;\n}\n\n\/* ===== Paragraphs ===== *\/\n.product-description.rte p {\n  margin-bottom: 20px;\n  line-height: 1.8;\n  color: #4a4f57;\n  font-size: 16px;\n}\n\n\/* ===== Paragraph Images (only in description) ===== *\/\n.product-description.rte img {\n  display: block;\n  max-width: 100%;\n  height: auto;\n  margin: 24px auto;\n  border-radius: 12px;\n  box-shadow: 0 8px 24px rgba(0, 0, 0, 0.06);\n}\n\n\/* ===== Lists ===== *\/\n.product-description.rte ul {\n  padding-left: 20px;\n  margin-bottom: 16px;\n  list-style-type: disc;\n}\n.product-description.rte li {\n  margin-bottom: 6px;\n}\n\n\/* ===== Links ===== *\/\n.product-description.rte a {\n  color: #0288D1;\n  text-decoration: none;\n  transition: color 0.3s ease;\n}\n.product-description.rte a:hover {\n  color: #01579B;\n}\n\n\/* ===== M5Stack-style Spec Table ===== *\/\n.product-description.rte .table-container {\n  width: 100%;\n  margin: 24px 0;\n  overflow-x: auto;\n}\n.product-description.rte .table-container table {\n  width: 100%;\n  border-collapse: collapse;\n  font-family: inherit;\n  font-size: 14px;\n  color: #2c2f36;\n  background-color: #ffffff;\n  border-radius: 10px;\n  overflow: hidden;\n  box-shadow: 0 4px 12px rgba(0, 0, 0, 0.04);\n}\n.product-description.rte .table-container thead th {\n  background: #1f6fe5;\n  color: #ffffff;\n  padding: 14px 16px;\n  text-align: center;\n  font-weight: 600;\n  border: none;\n  vertical-align: middle;\n}\n.product-description.rte .table-container td {\n  padding: 14px 16px;\n  border-bottom: 1px solid #e6ebf2;\n  vertical-align: middle;\n}\n.product-description.rte .table-container tbody td:first-child {\n  width: 36%;\n  text-align: center;\n  font-weight: 500;\n  color: #4a4f57;\n  background-color: rgba(31, 111, 229, 0.02);\n}\n.product-description.rte .table-container tbody tr:nth-child(even) {\n  background-color: #f7f9fc;\n}\n@media (hover: hover) {\n  .product-description.rte .table-container tbody tr:hover {\n    background-color: #eef3ff;\n  }\n}\n\n\/* ===== Learn \u0026 Documents Button ===== *\/\n.product-description.rte #learn-and-documents-btn {\n  display: inline-flex;\n  align-items: center;\n  padding: 6px 14px;\n  background-color: #1f6fe5;\n  color: #ffffff;\n  font-weight: 600;\n  border-radius: 6px;\n  margin-top: 8px;\n  text-decoration: none;\n  transition: background 0.3s ease;\n}\n.product-description.rte #learn-and-documents-btn:hover {\n  background-color: #1661c1;\n}\n\n\/* ===== Responsive Design for Smaller Screens (with table adjustments) ===== *\/\n@media screen and (max-width: 768px) {\n  .product-description.rte .table-container table,\n  .product-description.rte .table-container thead,\n  .product-description.rte .table-container tbody,\n  .product-description.rte .table-container th,\n  .product-description.rte .table-container td,\n  .product-description.rte .table-container tr {\n    display: block;\n    width: 100%;\n  }\n  .product-description.rte .table-container thead {\n    display: none;\n  }\n  .product-description.rte .table-container tbody tr {\n    padding: 12px 0;\n    border-bottom: 1px solid #e6ebf2;\n  }\n  .product-description.rte .table-container tbody td {\n    padding: 8px 16px;\n    border: none;\n    text-align: left;\n  }\n  .product-description.rte .table-container tbody td:first-child {\n    width: 100%;\n    background: none;\n    color: #1f6fe5;\n    font-weight: 600;\n    padding-top: 16px;\n  }\n  .product-description.rte .table-container tbody td:last-child {\n    padding-bottom: 16px;\n    color: #2c2f36;\n    line-height: 1.6;\n  }\n\n  \/* Ensure button appears correctly on mobile *\/\n  .product-description.rte #learn-and-documents-btn {\n    font-size: 14px;\n    padding: 8px 16px;\n  }\n}\n\u003c\/style\u003e\n\u003ch2 data-id=\"Description\" id=\"description\"\u003eDescription\u003c\/h2\u003e\n\u003cp\u003e\u003cstrong\u003eLLM-8850 Kit\u003c\/strong\u003e is a high-performance AI accelerator kit designed for edge AI and embedded computing scenarios. It consists of the \u003cstrong\u003eLLM-8850 Card\u003c\/strong\u003e AI accelerator and the \u003cstrong\u003eLLM-8850 PiHat\u003c\/strong\u003e adapter board. The core \u003cstrong\u003eLLM-8850 Card\u003c\/strong\u003e is based on the \u003cstrong\u003eAxera AX8850 SoC\u003c\/strong\u003e and adopts the \u003cstrong\u003eM.2 M-Key 2242\u003c\/strong\u003e form factor, delivering up to \u003cstrong\u003e24 TOPS @ INT8\u003c\/strong\u003e of computing power in a compact size. It efficiently accelerates multimodal large-model inference and video analytics workloads. The card integrates an \u003cstrong\u003eactive cooling system with a micro blower fan + CNC aluminum fin heatsink\u003c\/strong\u003e, with intelligent speed control driven by the onboard EC according to a temperature–current curve. This ensures stable low-temperature operation even under long-term full load or in enclosed chassis environments, preventing thermal throttling. The accompanying \u003cstrong\u003eLLM-8850 PiHat\u003c\/strong\u003e adapter board provides a \u003cstrong\u003e2.54-40P bus\u003c\/strong\u003e and a \u003cstrong\u003eUSB Type-C PD power input\u003c\/strong\u003e, and connects to the \u003cstrong\u003eRaspberry Pi 5\u003c\/strong\u003e via a \u003cstrong\u003ePCIe expansion FPC adapter\u003c\/strong\u003e, enabling seamless integration of high-performance AI acceleration on the Raspberry Pi platform for building compact and efficient edge intelligence systems.\u003c\/p\u003e\n\u003cp\u003e\u003cimg src=\"https:\/\/cdn.shopify.com\/s\/files\/1\/0056\/7689\/2250\/files\/DSC00401.webp?v=1770109614\u0026amp;width=700\" alt=\"LLM-8850 Kit\" style=\"width: 80%; height: auto;\"\u003e\u003c\/p\u003e\n\u003ch2 data-id=\"Features\" id=\"features\"\u003eFeatures\u003c\/h2\u003e\n\u003cul\u003e\n\u003cli\u003eUltra-compact form factor: M.2 M-Key 2242 size\u003c\/li\u003e\n\u003cli\u003eHigh-performance NPU: 24 TOPS @ INT8, octa-core Cortex-A55 1.7 GHz CPU\u003c\/li\u003e\n\u003cli\u003eIntelligent cooling \/ power: onboard blower fan + CNC aluminum integrated heatsink, EC-monitored temperature–current–speed closed loop\u003c\/li\u003e\n\u003cli\u003e8GB (4+4) high-bandwidth memory: 64-bit LPDDR4x, 4266 Mbps\u003c\/li\u003e\n\u003cli\u003eHardware video engine: 8K @ 30 fps H.264\/H.265 encoding, 8K @ 30 fps decoding, supports 16-channel 1080P parallel decoding\u003c\/li\u003e\n\u003cli\u003eSecure boot \u0026amp; encryption: AES \/ DES \/ 3DES \/ SHA-256 hardware security module\u003c\/li\u003e\n\u003cli\u003eNative AXCL support: one-click full-stack model deployment for CNN, Transformer, CLIP, Whisper, Llama3.2, Qwen3, InternVL3, etc., with H.264\/H.265 simultaneous encode\/decode transcoding\u003c\/li\u003e\n\u003cli\u003eUSB Type-C PD power input:\n\u003cul\u003e\n\u003cli\u003eUSB PD3.0 100W protocol\u003c\/li\u003e\n\u003cli\u003eDC 24V overvoltage protection\u003c\/li\u003e\n\u003cli\u003ePower capability requirement \u0026gt; 9V@3A (27W)\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ch2 data-id=\"Includes\" id=\"includes\"\u003eIncludes\u003c\/h2\u003e\n\u003ch3 data-id=\"LLM-8850 Kit (8G) (SKU:AI-002-8G)\" id=\"llm-8850 kit (8g) (sku:ai-002-8g)\"\u003eLLM-8850 Kit (8G) (SKU:AI-002-8G)\u003c\/h3\u003e\n\u003cul\u003e\n\u003cli\u003e1 x LLM-8850 Card (8G)\u003c\/li\u003e\n\u003cli\u003e1 x LLM-8850 PiHat\u003c\/li\u003e\n\u003cli\u003e1 x 2.54-40P Female Header\u003c\/li\u003e\n\u003cli\u003e1 x Screw and Standoff Accessory Pack\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cimg src=\"https:\/\/cdn.shopify.com\/s\/files\/1\/0056\/7689\/2250\/files\/7_ce065942-3ebd-450f-9a6e-0e6a4e26e596.webp?v=1770109639\u0026amp;width=700\" alt=\"LLM-8850 Kit\" style=\"width: 70%; height: auto;\"\u003e\u003c\/p\u003e\n\u003ch2 data-id=\"Applications\" id=\"applications\"\u003eApplications\u003c\/h2\u003e\n\u003cul\u003e\n\u003cli\u003eEdge AI inference deployment\u003c\/li\u003e\n\u003cli\u003eEmbedded video analytics\u003c\/li\u003e\n\u003cli\u003eRaspberry Pi AI computing expansion\u003c\/li\u003e\n\u003cli\u003eMultimodal large-model implementation\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cimg src=\"https:\/\/cdn.shopify.com\/s\/files\/1\/0056\/7689\/2250\/files\/6_1123a8a5-9832-4371-b31f-b5ac20508668.webp?v=1770109660\u0026amp;width=700\" alt=\"LLM-8850 Kit\" style=\"width: 80%; height: auto;\"\u003e\u003c\/p\u003e\n\u003ch2 data-id=\"Specifications\" id=\"specifications\"\u003eSpecifications\u003c\/h2\u003e\n\u003cdiv class=\"table-container\"\u003e\n\u003ctable\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth\u003eSpecification\u003c\/th\u003e\n\u003cth\u003eParameter\u003c\/th\u003e\n\u003c\/tr\u003e\n\u003c\/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003eSoC\u003c\/td\u003e\n\u003ctd\u003eAxera AX8850\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCPU\u003c\/td\u003e\n\u003ctd\u003eOcta-core Cortex-A55 1.7 GHz\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eNPU\u003c\/td\u003e\n\u003ctd\u003e24 TOPS @ INT8\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eVideo Encoder\u003c\/td\u003e\n\u003ctd\u003e8K @ 30 fps H.264\/H.265 encoding, supports scaling \/ cropping\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eVideo Decoder\u003c\/td\u003e\n\u003ctd\u003e8K @ 30 fps H.264\/H.265 decoding, supports 16-channel 1080P parallel decoding, scaling \/ cropping\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMemory\u003c\/td\u003e\n\u003ctd\u003eOptional: 8GB 64-bit LPDDR4x, 4266 Mbps (4+4) \/ 4GB 64-bit LPDDR4x, 4266 Mbps (2+2)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eStorage\u003c\/td\u003e\n\u003ctd\u003e32Mbits QSPI NOR Flash (bootloader only)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eForm Factor\u003c\/td\u003e\n\u003ctd\u003eM.2 M-Key 2242, PCIe 2.0 x2\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCooling\u003c\/td\u003e\n\u003ctd\u003eMicro blower fan + CNC aluminum integrated heatsink, EC intelligent thermal control\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eOperating Temperature\u003c\/td\u003e\n\u003ctd\u003e0 ~ 60 °C\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eFull-load Temperature at Room Temp\u003c\/td\u003e\n\u003ctd\u003e70 °C\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eAdapter Board PD Power\u003c\/td\u003e\n\u003ctd\u003eUSB PD3.0 100W protocol (DC 9V \/ 12V \/ 20V)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eAdapter Board Load Capability\u003c\/td\u003e\n\u003ctd\u003eOutput to Raspberry Pi 5: DC 5V@4A; Output to M.2 accelerator card: DC 3.3V@6A\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eLLM-8850 Card Power\u003c\/td\u003e\n\u003ctd\u003e7W @ 3.3V\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePower Requirement\u003c\/td\u003e\n\u003ctd\u003e\u0026gt; 9V@3A (27W)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eProduct Size\u003c\/td\u003e\n\u003ctd\u003eLLM-8850 Card: 42.6 x 24.0 x 9.7mm \u003cbr\u003eLLM-8850 PiHat: 65.0 x 58.0 x 12.7mm\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eProduct Weight\u003c\/td\u003e\n\u003ctd\u003e31.9g\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePackage Size\u003c\/td\u003e\n\u003ctd\u003e165.0 x 90.0 x 17.5mm\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eGross Weight\u003c\/td\u003e\n\u003ctd\u003e85.0g\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\n\u003cp\u003e\u003cimg alt=\"LLM-8850 Kit\" style=\"width: 80%; height: auto;\" src=\"https:\/\/cdn.shopify.com\/s\/files\/1\/0056\/7689\/2250\/files\/1920x2260_8e35dc71-26b1-4ba3-a719-3efccf277e66.webp?v=1770109708\u0026amp;width=700\"\u003e\u003c\/p\u003e\n\u003cdiv style=\"background: #ffebee; border: 1px solid #ffccd1; border-left: 4px solid #d32f2f; border-radius: 8px; padding: 14px 16px; margin: 24px 0;\"\u003e\n\u003cstrong style=\"display: block; color: #b71c1c; font-size: 15px; margin-bottom: 6px;\"\u003e Power Supply Notice \u003c\/strong\u003e \u003cspan style=\"color: #2c2c2c; font-size: 14.5px; line-height: 1.6;\"\u003e Please power the entire system via the USB Type-C PD input on the adapter board. The power supply must provide at least 9V@3A (27W). Do not supply power from the Raspberry Pi 5 side to avoid device damage. \u003c\/span\u003e\n\u003c\/div\u003e\n\u003cdiv style=\"background: #fff8e1; border: 1px solid #ffe0a3; border-left: 4px solid #f9a825; border-radius: 8px; padding: 14px 16px; margin: 24px 0;\"\u003e\n\u003cstrong style=\"display: block; color: #8d6e00; font-size: 15px; margin-bottom: 6px;\"\u003e Reminder \u003c\/strong\u003e \u003cspan style=\"color: #2c2c2c; font-size: 14.5px; line-height: 1.6;\"\u003e The device may become hot during operation. Do not touch it to avoid burns. \u003c\/span\u003e\n\u003c\/div\u003e\n\u003cdiv style=\"background: #fff8e1; border: 1px solid #ffe0a3; border-left: 4px solid #f9a825; border-radius: 8px; padding: 14px 16px; margin: 24px 0;\"\u003e\n\u003cstrong style=\"display: block; color: #8d6e00; font-size: 15px; margin-bottom: 6px;\"\u003e Device Requirement \u003c\/strong\u003e \u003cspan style=\"color: #2c2c2c; font-size: 14.5px; line-height: 1.6;\"\u003e Due to limited interrupt resources inside the Raspberry Pi, when using PCIe to a 4-lane M.2 adapter, M.2 SSDs and the LLM-8850 cannot be used simultaneously. \u003c\/span\u003e\n\u003c\/div\u003e\n\u003ch2\u003eLearn and Documents\u003c\/h2\u003e\n → DOCUMENTS \n\u003c\/div\u003e","brand":"m5stack-store","offers":[{"title":"Default Title","offer_id":49566419812579,"sku":"grf1787247290074","price":150.5,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0734\/2961\/4819\/files\/ai-8850-llm-accelerator-m2-kit-8gb-version-ax8850_1.webp?v=1787247296","url":"https:\/\/regalcladfashion.shop\/products\/ai-8850-llm-accelerator-m2-kit-8gb-version-ax8850","provider":"My Store","version":"1.0","type":"link"}