{"product_id":"100gbe-qsfp28-ethernet-switch-used-for-mellanox-sn3800-cs2f","title":"100GbE QSFP28 Ethernet Switch Used for Mellanox SN3800-CS2F 64-Port 2U Network Switch","description":"\u003ch2\u003eNVIDIA Mellanox Spectrum-3 SN3800 Multi-Speed Datacenter Switch\u003c\/h2\u003e\n\u003cp\u003eThe \u003cstrong\u003eNVIDIA Mellanox Spectrum-3 SN3800 Multi-Speed Datacenter Switch\u003c\/strong\u003e is a high-performance datacenter Ethernet switch powered by the NVIDIA Spectrum-3 ASIC, delivering 12.8 Tbps full duplex of non-blocking throughput with Variable: up to 36×200GbE \/ 64×100GbE \/ 128×50GbE \/ 256×25GbE \/ 8×400GbE ports. Designed for 400G\/200G cloud spine, it provides \u0026lt; 250 ns cut-through cut-through latency and 64 MB shared packet buffer shared packet buffer for lossless RoCE v2 RDMA transport.\u003c\/p\u003e\n\u003cp\u003e\u003cem\u003eCondition: Refurbished \/ Pre-owned — fully tested, cleaned, and verified. Ships with warranty coverage.\u003c\/em\u003e\u003c\/p\u003e\n\u003ctable style=\"width:100%;border-collapse:collapse;margin:12px 0;font-size:14px;\"\u003e\\\n\u003cthead\u003e\u003ctr\u003e\n\u003cth style=\"background:#f8f9fa;padding:8px;border:1px solid #dee2e6;width:35%;\"\u003eSpecification\u003c\/th\u003e\n\u003cth style=\"background:#f8f9fa;padding:8px;border:1px solid #dee2e6;\"\u003eDetails\u003c\/th\u003e\n\u003c\/tr\u003e\u003c\/thead\u003e\\\n\u003ctbody\u003e\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eChip\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eSpectrum-3\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003ePorts\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eVariable: up to 36×200GbE \/ 64×100GbE \/ 128×50GbE \/ 256×25GbE \/ 8×400GbE\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eThroughput\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e12.8 Tbps full duplex\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eLatency\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e\u0026lt; 250 ns cut-through\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003ePacket Rate\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e9.52 Bpps\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eBuffer\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e64 MB shared packet buffer\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003ePower\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e≤ 350W typical\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eDimensions\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e440 × 470 × 43 mm (1U)\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eWeight\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e8.5 kg\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eInterfaces\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e1× RJ45 management, 1× micro-USB console, 1× USB 3.0\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eOs\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eCumulus Linux, SONiC, MLNX-OS\u003c\/td\u003e\n\u003c\/tr\u003e\\\\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eFeatures\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eVXLAN, EVPN, RoCE v2, PFC, ECN, DCB, NAT, VRF, ACL 24K, Advanced Mirroring, sFlow, QoS per-queue\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\n\u003ch3\u003eKey Features\u003c\/h3\u003e\u003cul\u003e\n\u003cli\u003eVXLAN\u003c\/li\u003e\n\u003cli\u003eEVPN\u003c\/li\u003e\n\u003cli\u003eRoCE v2\u003c\/li\u003e\n\u003cli\u003ePFC\u003c\/li\u003e\n\u003cli\u003eECN\u003c\/li\u003e\n\u003cli\u003eDCB\u003c\/li\u003e\n\u003cli\u003eNAT\u003c\/li\u003e\n\u003cli\u003eVRF\u003c\/li\u003e\n\u003cli\u003eACL 24K\u003c\/li\u003e\n\u003cli\u003eAdvanced Mirroring\u003c\/li\u003e\n\u003cli\u003esFlow\u003c\/li\u003e\n\u003cli\u003eQoS per-queue\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003ch3\u003eTypical Applications\u003c\/h3\u003e\u003cp\u003e400G\/200G cloud spine, AI\/ML large-scale training, HPC, storage NVMe-oF, telecom core\u003c\/p\u003e\n\u003ch3\u003eFrequently Asked Questions\u003c\/h3\u003e\\\n\u003cdetails style=\"margin:8px 0;padding:8px;border:1px solid #e0e0e0;border-radius:4px;\"\u003e\u003csummary style=\"cursor:pointer;font-weight:600;color:#1a1a1a;\"\u003eWhat speeds does the SN3800 support?\u003c\/summary\u003e\u003cp style=\"margin:8px 0 0;color:#444;\"\u003eThe SN3800 supports 25G, 50G, 100G, 200G, and 400GbE simultaneously on different ports—making it ideal for mixed-speed cloud spine deployments.\u003c\/p\u003e\u003c\/details\u003e\\\\n\u003cdetails style=\"margin:8px 0;padding:8px;border:1px solid #e0e0e0;border-radius:4px;\"\u003e\u003csummary style=\"cursor:pointer;font-weight:600;color:#1a1a1a;\"\u003eHow does SN3800 compare to SN3700?\u003c\/summary\u003e\u003cp style=\"margin:8px 0 0;color:#444;\"\u003eSN3800 uses the Spectrum-3 ASIC with 12.8 Tbps throughput (vs 7.2 Tbps), 64 MB buffer (vs 16 MB), and supports 400G ports. It's designed for next-generation 400G spine while SN3700 targets 100G\/200G leaf.\u003c\/p\u003e\u003c\/details\u003e\\\\n\u003cdetails style=\"margin:8px 0;padding:8px;border:1px solid #e0e0e0;border-radius:4px;\"\u003e\u003csummary style=\"cursor:pointer;font-weight:600;color:#1a1a1a;\"\u003eCan SN3800 be used as a spine switch?\u003c\/summary\u003e\u003cp style=\"margin:8px 0 0;color:#444;\"\u003eYes. With up to 8×400GbE or 36×200GbE ports and 12.8 Tbps throughput, the SN3800 is designed for high-density cloud spine layers.\u003c\/p\u003e\u003c\/details\u003e\\\\n\u003cdetails style=\"margin:8px 0;padding:8px;border:1px solid #e0e0e0;border-radius:4px;\"\u003e\u003csummary style=\"cursor:pointer;font-weight:600;color:#1a1a1a;\"\u003eWhat buffer size does SN3800 have?\u003c\/summary\u003e\u003cp style=\"margin:8px 0 0;color:#444;\"\u003e64 MB shared packet buffer—4× larger than SN3700—ensuring zero packet loss even under heavy RoCE v2 RDMA traffic in AI training clusters.\u003c\/p\u003e\u003c\/details\u003e\n\u003cp\u003e\u003cem\u003eAvailable at \u003ca href=\"https:\/\/zhongxindata.hk\"\u003eZhongxin Data\u003c\/a\u003e — enterprise network hardware specialists. Worldwide shipping from Hong Kong.\u003c\/em\u003e\u003c\/p\u003e","brand":"NVIDIA Mellanox","offers":[{"title":"Default Title","offer_id":55807042748708,"sku":"ALI-1601824487232","price":5610.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1017\/6223\/6708\/files\/100GbE-QSFP28-Ethernet-Switch-Used-for_21422771-da21-4fb4-b909-ad3a4d06a310.webp?v=1784864303","url":"https:\/\/zhongxindata.hk\/products\/100gbe-qsfp28-ethernet-switch-used-for-mellanox-sn3800-cs2f","provider":"Zhongxin Data","version":"1.0","type":"link"}