Product Brief
Intel® Ethernet Server
Adapter XL710 for OCP
Industry-leading, energy-ecient design for
10/40GbE performance and multi-core processors.
Key Features
• OCP Spec. v2.0, Type 1
• Supports 4x10GbE, 1x40GbE or
2x40GbE configurations¹
• PCI Express* (PCIe) v3.0, x8
• Exceptional low power adapters
• Network Virtualization offloads
including VXLAN, GENEVE,
NVGRE, MPLS, and VXLAN-GPE
with Network Service
Headers (NSH)
• Intel® Ethernet Flow Director for
hardware based application
traffic steering
• Dynamic Device Personalization
(DDP) enables increased packet
processing efficiency for NFV and
Cloud deployments
• Data Plane Development Kit
(DPDK) optimized for efficient
packet processing
• Excellent small packet
performance for network²
appliances and Network Functions
Virtualization (NFV)
• Intelligent offloads to enable high
performance on servers with
Intel® Xeon® processors
• I/O virtualization innovations for
maximum performance in a
virtualized server
Overview
As a founding member of OCP, Intel strives to increase the number of open
solutions based on OCP specications. The Intel® Ethernet Network
Adapter XL710 for OCP is part of the Intel® Ethernet 700 Series, and oers
10/40GbE port speeds.
Intel® Ethernet 700 Series Network Adapters are the foundation for server
connectivity, providing broad interoperability, critical performance
optimizations, and increased agility for Communications, Cloud, and
Enterprise IT network solutions.
• Interoperability - Multiple speeds and media types for broad compatibility
backed by extensive testing and validation.
• Optimization - Intelligent ooads and accelerators to unlock network
performance in servers with Intel® Xeon® processors.
• Agility - Both Kernel and Data Plane Development Kit (DPDK) drivers for
scalable packet processing.
Built on more than 35 years of continuous Ethernet innovations, the
Intel® Ethernet 700 Series delivers networking performance across a wide
range of network port speeds through intelligent ooads, sophisticated
packet processing, and quality open source drivers.
All Intel® Ethernet 700 Series Network Adapters
include these feature-rich technologies:
Flexible and Scalable I/O for Virtualized
Infrastructures
Intel® Virtualization Technology (Intel® VT), delivers
outstanding I/O performance in virtualized server
environments.
I/O bottlenecks are reduced through intelligent ooads
such as Virtual Machine Device Queues (VMDq) and
Flexible Port Partitioning, using SR-IOV with a common
Virtual Function driver for networking trac per Virtual
Machine (VM), enabling near-native performance and
VM scalability. Host-based features supported include:
VMDq for Emulated Path: VMDq, enables a hypervisor
to represent a single network port as multiple network
ports that can be assigned to the individual VMs.
Trac handling is ooaded to the network controller,
delivering the benets of port partitioning with little to
no administrative overhead by the IT sta.
SR-IOV for Direct Assignment: Adapter-based
isolation and switching for various virtual station
instances enables optimal CPU usage in virtualized
environments.
• Up to 128 VFs, each VF can support a unique and
separate data path for I/O related functions within the
PCI Express* hierarchy.
• Use of SR-IOV with a networking device, for example,
allows the bandwidth of a single port (function) to be
partitioned into smaller slices that can be allocated to
specific VMs or guests, via a standard interface.
Intel® Ethernet Adaptive Virtual Function (Intel®
Ethernet AVF): A virtual function driver that eases
SR-IOV hardware upgrades or changes, and preserves
base mode functionality in hardware and software.
Customers deploying mass-scale VMs or containers
for their network infrastructure now have a common
Virtual Function (VF) driver. This driver supports a
Base Mode and advanced set of features on the Intel®
Ethernet 700 Series.
Enhanced Network Virtualization Overlays (NVO)
Network virtualization has changed the way networking
is done in the data center, delivering accelerations
across a wide range of tunneling methods.
VXLAN, GENEVE, NVGRE, MPLS, and VXLAN-GPE
with NSH Ooads: These stateless ooads preserve
application performance for overlay networks, and the
network trac can be distributed across CPU cores,
increasing network throughput.
Flexible Port Partitioning (FPP)
FPP leverages the PCI-SIG SR-IOV specication. Virtual
controllers can be used by the Linux* host directly and/
or assigned to virtual machines.
• Assign up to 63 Linux host processes or virtual
machines per port to virtual functions.
• Control the partitioning of per port bandwidth across
multiple dedicated network resources, ensuring
balanced QoS by giving each assigned virtual controller
equal access to the ports bandwidth.
Network administrators can also rate limit each of
these services to control how much of the pipe is
available to each process.
Greater Intelligence and Performance for NFV and
Cloud deployments
Dynamic Device Personalization (DDP) customizable
packet filtering, along with enhanced Data Plane
Development Kit (DPDK), support advanced packet
forwarding and highly-efficient packet processing for
both Cloud and Network Functions Virtualization
(NFV) workloads.
• DDP enables workload-specific optimizations, using
the programmable packet-processing pipeline.
Additional protocols can be added to the default set
to improve packet processing efficiency that results
in higher throughput and reduced latency. With the
700 Series, new protocols can be added or modified
on-demand, and applied at run-time using Software
Defined Firmware or APIs, eliminating the need to reset
or reboot the server. This not only keeps the server
and VMs up, running, and computing, it also increases
performance for VNFs that process network traffic that
is not included in the default firmware.
Download DDP Profiles
• DPDK provides a programming framework for Intel®
processors and enables faster development of highspeed data packet networking applications.
Advanced Traffic Steering
Intel® Ethernet Flow Director (Intel® Ethernet FD) is
an advanced trac steering capability built into the
adapter. It consists of a large number of ow anity
lters that direct receive packets by their ows to
queues for classication, load balancing, and matching
between ows and CPU cores.
Steering trac into specic queues can eliminate
context switching required within the CPU. As a result,
Intel® Ethernet FD signicantly increases the number of
transactions per second and reduces latency for cloud
applications like memcached.
2
2