With the evolution of data center virtualization technology and the popularization of distributed computing, the internal traffic of the data center has increased exponentially. The current traditional 10G/100G architecture cannot carry the ever-increasing business demands and more east-west characteristic traffic.
The 25G/100G data center network adopts the large two-layer structure of IP CLOS to cope with 80% of the east-west traffic of the cloud computing data center. The server interaction delay is smaller and the overlay network is constructed with higher flexibility.
Flexible 10/25/40/100GbE Interface Speeds, Support Stacking
Support MLAG, BGPv4/v6, EVPN-VXLAN, REUP, GR, BFD
Support PFC, ECN, RDMA over Converged Ethernet (RoCE)
Flexible 10/25/40/50/100GbE Interface Speeds, Support Stacking
Support MLAG, BGPv4/v6, EVPN-VXLAN, REUP, GR, BFD
Support PFC, ECN, RDMA over Converged Ethernet (RoCE)
Industry Leading Low Latency RoCE
Support Windows, RHEL/CentOS, VMware, FreeBSD, OFED and WinOF-2
Dual 25G Ports Apply to 25G/10G/1G Transceivers, DACs and AOCs
Built-in Security, and Remote Management Capabilities with Intel Xeon Processor
DDR4 SDRAM Slot - Reduces Processor's Idle Time and Enhances Performance
Optimize Business Network's Performance with 10 Gigabit Ethernet Connectivity
< 1W Low Power Consumption
LC Duplex Type, VCSEL+PIN
Fully Compatible with 200+ Vendors
High Quality MACOM Chip
VCSEL+PIN, Built-in Dual CDR
≤1.8W (Typical) Low Power Consumption
Fully Compatible with All FHD® Enclosures
Tool-less Snap Installation for Easier MACs
Clear Numbering for Quick Fiber Identification
Fully Compatible with All FHD® Enclosures
Tool-less Snap Installation for Easier MACs
Clear Numbering for Quick Fiber Identification
Constructed in units of POD, the data center can expand with the business volume.
With a reasonable convergence ratio, the expansion of LEAF nodes does not require network redesign.
Through SPAN, RSPAN, ERSPAN to achieve data drainage, traffic analysis and traffic cleaning.
Automated network deployment and unified monitoring management are realized through NETCONF, OpenFlow, Python, gRPC, Sflow, SNMP, etc.
Complete QOS strategy and queue scheduling mechanism to ensure congestion forwarding.
An end-to-end, lossless, and low-latency forwarding RDMA basic bearer network can be constructed.
End Chat?
Do you really want to close the chat?