[{"data":1,"prerenderedAt":10},["ShallowReactive",2],{"tech-en-Sopnon-Series\u002Fsopnon-series":3},{"slug":4,"title":5,"seoTitle":6,"description":5,"keywords":5,"hero":7,"html":8,"related":9},"Sopnon-Series\u002Fsopnon-series","Sopnon Series","Sopnon Series Page","https:\u002F\u002Fditlcm.oss-cn-beijing.aliyuncs.com\u002Fimages\u002Ftech\u002Fbanner_SOPNON.jpg","\u003Cp>\u003Cimg alt=\"\" height=\"785\" src=\"https:\u002F\u002Fditlcm.oss-cn-beijing.aliyuncs.com\u002Fimages\u002Ftech\u002Fbanner_SOPNON.jpg\"\r\nwidth=\"1920\">\u003C\u002Fp>\r\n\r\n\r\n\r\n\u003Cp class=\"box-product__text\" style=\"text-align: left;\">\r\n\u003Cspan class=\"box-product__text__fontWeight700\">SOPNON \u003C\u002Fspan> \r\nrevolutionizes \r\n\u003Cspan class=\"box-product__text__fontWeight700\">AI computing \u003C\u002Fspan> \r\nfrom \r\n\u003Cspan class=\"box-product__text__fontWeight700\">cloud to edge \u003C\u002Fspan>\r\nwith its \r\n\u003Cspan class=\"box-product__text__fontWeight700\">proprietary TPU architecture \u003C\u002Fspan>\r\n. Our solutions deliver \r\n\u003Cspan class=\"box-product__text__fontWeight700\">exceptional energy efficiency \u003C\u002Fspan>\r\nand \r\n\u003Cspan class=\"box-product__text__fontWeight700\">high-performance processing \u003C\u002Fspan>\r\n, making us a leader in \r\n\u003Cspan class=\"box-product__text__fontWeight700\">AI-driven applications \u003C\u002Fspan>\r\nacross industries like \r\n\u003Cspan class=\"box-product__text__fontWeight700\">robotics \u003C\u002Fspan>, \r\n\u003Cspan class=\"box-product__text__fontWeight700\">automotive \u003C\u002Fspan>, \r\nand\r\n\u003Cspan class=\"box-product__text__fontWeight700\">AIoT\u003C\u002Fspan>.\r\n\u003C\u002Fp>\r\n\r\n\u003Cp style=\"text-align: left;\">\u003Cstrong>\u003Cspan style=\"color:#28804a; text-align: left;\">Core Products:\u003C\u002Fspan>\u003C\u002Fstrong>\u003C\u002Fp>\r\n\u003Cul style=\"text-align: left;\">\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nIndustrial Motherboards:\r\n\u003C\u002Fspan>\r\nARM-based embedded motherboards designed for customization, reliability, and long-lasting performance in industrial environments.\r\n\u003C\u002Fli>\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nIndustrial Display Modules: \r\n\u003C\u002Fspan>\r\nHigh-brightness LCDs with capacitive touchscreens (5-43 inches), built for demanding industrial environments. \r\n\u003C\u002Fli>\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nAll-in-One Industrial Computers: \r\n\u003C\u002Fspan>\r\n24+ interfaces with flexible installation options for seamless integration in complex applications. \r\n\u003C\u002Fli>\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nIndustrial Mini Pcs: \r\n\u003C\u002Fspan>\r\nRuggedized, dust-resistant, and EMI-proof for plug-and-play functionality in harsh environments.\r\n\u003C\u002Fli>\r\n\u003C\u002Ful>\r\n\r\n\u003Cp style=\"text-align: left;\">\u003Cstrong>\u003Cspan style=\"color:#28804a; text-align: left;\">Performance Features:\u003C\u002Fspan>\u003C\u002Fstrong>\u003C\u002Fp>\r\n\u003Cul style=\"text-align: left;\">\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nTPU Systolic Array: \r\n\u003C\u002Fspan>\r\nOptimized for INT8\u002FFP16 mixed-precision with 85% hardware utilization, accelerating AI tasks.\r\n\u003C\u002Fli>\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nHigh Compute Density: \r\n\u003C\u002Fspan>\r\n16 TOPS at INT8 precision for edge chips and 400+ TFLOPS for cloud chips, with 32MB on-chip SRAM to reduce latency. \r\n\u003C\u002Fli>\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nVideo-AI Co-Processing: \r\n\u003C\u002Fspan>\r\nSimultaneous video decoding and AI inference for up to 100 video streams.\r\n\u003C\u002Fli>\r\n\u003Cli>\r\n\u003Cspan class=\"box-product__label\">\r\nFull-Stack Toolchain: \r\n\u003C\u002Fspan>\r\nUnified SDK (OKNN) for seamless deployment across TensorFlow and PyTorch, cutting algorithm migration costs.\r\n\u003C\u002Fli>\r\n\u003C\u002Ful>\r\n\r\n\r\n\u003Cp style=\"text-align: left;\">\u003Cstrong>\u003Cspan style=\"color:#28804a; text-align: left;\">Specifications:\u003C\u002Fspan>\u003C\u002Fstrong>\u003C\u002Fp>\r\n\r\n\u003Ctable border=\"0\" cellpadding=\"1\" cellspacing=\"1\" class=\"mod-table table-list-view has-wrapper\"\r\nstyle=\"width:100%;\">\r\n\u003Ctbody>\r\n\u003Ctr>\r\n\u003Ctd rowspan=\"7\" style=\"width: 256px;\">\u003Cimg alt=\"\" height=\"236\"\r\nsrc=\"\u002Fimages\u002Ftech\u002FBM1688.jpg\" style=\"float:left\"\r\nwidth=\"190\">\u003C\u002Ftd>\r\n\u003Ctd style=\"width: 121px; text-align: left; background-color: rgb(244, 248, 251);\">\r\nCPU\r\n\u003C\u002Ftd>\r\n\u003Ctd style=\"width: 814px; text-align: left; background-color: rgb(244, 248, 251);\">\r\nBM1688 Octa-core Cortex-A53\r\n\u003C\u002Ftd>\r\n\u003C\u002Ftr>\r\n\u003Ctr>\r\n\u003Ctd style=\"width: 121px; text-align: left;\">\r\nGPU\r\n\u003C\u002Ftd>\r\n\u003Ctd style=\"width: 814px; text-align: left;\">\r\nNo discrete GPU; relies on TPU acceleration\r\n\u003C\u002Ftd>\r\n\u003C\u002Ftr>\r\n\u003Ctr>\r\n\u003Ctd style=\"width: 121px; text-align: left; background-color: rgb(244, 248, 251);\">\r\nNPU\r\n\u003C\u002Ftd>\r\n\u003Ctd style=\"width: 814px; text-align: left; background-color: rgb(244, 248, 251);\">\r\n16 TOPS at INT8 precision (supports INT4\u002FFP16 mixed-precision)\r\n\u003C\u002Ftd>\r\n\u003C\u002Ftr>\r\n\u003Ctr>\r\n\u003Ctd style=\"width: 121px; text-align: left; \">\r\nStorage\r\n\u003C\u002Ftd>\r\n\u003Ctd style=\"width: 814px; text-align: left;\">\r\nDDR: 4GB\u002F8GB\u002F16GB + eMMC: 32GB\u002F64GB\u002F128GB\r\n\u003C\u002Ftd>\r\n\u003C\u002Ftr>\r\n\u003Ctr>\r\n\u003Ctd colspan=\"1\" rowspan=\"2\" style=\"width: 121px; text-align: left; background-color: rgb(244, 248, 251);\">\r\nMultimedia\r\n\u003C\u002Ftd>\r\n\u003Ctd colspan=\"1\" rowspan=\"2\" style=\"width: 814px; text-align: left; background-color: rgb(244, 248, 251);\">\r\n16-channel 1080P decoding + 10-channel encoding, multi-channel AI analysis\r\n\u003C\u002Ftd>\r\n\u003C\u002Ftr>\r\n\u003C\u002Ftbody>\r\n\u003C\u002Ftable>",[],1789726565936]