Cisco Cisco UCS C22 M3 Rack Server Ficha De Dados

Página de 6
 
 
 
 
 
 
 
re 
ning
 
 
 
 
Product Brief
 
Intel® Ethernet Converged Network Adapters X710 
With Support for SFP+ Connections
 
Network Connectivity
 
 
 
 
 
Intel
®   
Ethernet Converged
 
Network Adapters 
X710 10 GbE
 
 
 
 
 
 
 
Extending Intel® Virtualization Technology beyon 
Server Virtualization to the Network with Hardwa 
Optimizations and Offloads for the Rapid Provisio 
of Networks in an Agile Data Center
 
 
Key Features
 
• Dual- and quad-port 10 GbE 
adapters
 
• PCI Express* (PCIe) 3.0, x8
 
• Exceptional Low Power Adapters
 
• Network Virtualization offloads 
including VXLAN, NVGRE, Geneve, 
Network Service Headers  (NSH)
2
 
• Intel® Ethernet Flow Director 
for hardware based application 
traffic steering
 
• Data Plane Developer Kit (DPDK) 
optimized for efficient packet 
processing
 
• Excellent small packet performance 
for network appliances and Network 
Function Virtualization (NFV)
 
• Intelligent offloads to enable high 
performance with Intel® Xeon® 
processor-based servers
 
• I/O virtualization innovations for 
maximum performance in a 
virtualized server
 
• Unified networking providing a single 
wire support for LAN and storage: 
NAS (SMB, NFS) and SAN (iSCSI)
1
 
Overview
 
Intel continues its legacy of Ethernet leadership by introducing a 10/40 
gigabit family of adapters powered by the Intel® Ethernet X710 Controller 
code-named Fortville.
 
The X710 adapter family addresses the demanding needs of the next-gener- 
ation agile data center by providing unmatched features for both server and 
network virtualization, flexibility for LAN and SAN networks and proven, reli- 
able performance.
 
 
Leading 10/40 GbE Performance
 
The X710 adapter family delivers superior performance with a theoretical 
throughput of 80 Gb/s (40Gb/s Tx; 40Gb/s Rx) bidirectional throughput (quad- 
port adapter required), in a PCI Express v3.0 x8 slot. Optimized performance 
vectors and key uses include:
 
• Small Packet Performance: Achieves wire-rate throughput on smaller pay- 
load sizes (>128 Bytes for 40 GbE and >64 Bytes for 10 GbE
 
• Bulk Transfer Performance: Delivers line-rate performance with low CPU 
usage for large application  buffers
 
• Virtualized Performance: Alleviates hypervisor I/O bottlenecks by providing 
flow separation for Virtual Machines  (VMs)
 
• Network Virtualization: Network virtualization overlay offloads including 
VXLAN, NVGRE, Geneve, Network Service Headers  (NSH)
2