Communications and networking equipment typically employs multi-architecture designs where control and application processing run on general purpose processors, while packet and signal processing execute on specialized processors. This multi-architecture approach has its advantages, but it leads to complex, inflexible designs. Today an alternative approach where applications, services, control plane, and packet processing are consolidated onto one hardware platform is possible thanks to performance gains in the latest multi-core Intel® Xeon processors. This 4:1 consolidation allows original equipment providers (OEMs) to streamline development, and offer service providers ways to reduce capital and operating expenditures (CapEx and OpEx). What’s more, Intel Xeon processors provide the scalability and flexibility to offer these benefits to a wide variety of markets, such as unified threat management (UTM), enterprise PBX systems, and LTE infrastructure.


How are these benefits possible? Let's start with the hardware. Consolidating workloads onto one architecture reduces the development, integration and validation issues that come from using multiple architectures. Using a single architecture also means fewer boards to build, inventory, maintain, and support. From a software perspective, workload consolidation means one code base and one set of development tools.  This reduces development costs and enables more software reuse, which in turn boosts efficiency, reduces training time and decreases license fees. Entire product lines can now scale from the low end of the bandwidth spectrum all the way up to high volume boxes with a single development effort.


A key challenge in workload consolidation is the fact that different workloads may require different operating systems.  This challenge can be overcome with virtualization, a technology that introduces a software layer known as a hypervisor below the OS level (Figure 1).  The hypervisor enables operation of multiple virtual machines (VMs), each containing a guest OS and its associated applications, by presenting each guest OS with what appears to be a dedicated hardware platform. Intel Xeon processors support virtualization through a hardware-assist technology called Intel® Virtualization Technology (Intel® VT) that performs a number of virtualization tasks directly in hardware, reducing the footprint of the hypervisor and improving performance.


norco figure 1-alternative.png

Figure 1. Embedded hypervisors enable multiple OSs to run on the same hardware.


The need for extreme packet and security performance has also hindered workload consolidation in the past.  Today, manufacturers can leverage the Intel® Data Plane Development Kit (Intel® DPDK) to achieve the packet performance they need and speed solutions to market. To speed up security functions, Intel Xeon processors also include optimizations for cryptography, data compression, and pattern matching. Intel® Streaming SIMD Extensions 4.2 (Intel® SSE 4.2) and Intel® AES New Instructions (Intel® AES-NI) set extensions accelerate tokenizing, regular expression evaluation, virus scanning, intrusion, and other security functions. OEMs can tap Intel® QuickAssist Technology to get workload-specific APIs for tasks such as pattern matching, cryptography, and data compression.

Let’s now look at how some Alliance members are using Intel Xeon processors to enable workload consolidation.


UTM is a fast-growing segment within the network security market that Advantech is addressing with a variety of Intel® Xeon® processor-based solutions. One in particular is the FWA-6500 Network Application Platform. The FWA-6500 provides a complete network communication appliance in a 2U form factor. It features the Intel® Xeon® processor 5500 or 5600 series with up to in 12 cores in a dual-socket configuration and up to 96 GB of DDR3 memory with ECC. The FWA-6500 is configurable with two PCI Express full-height, half-length add-on cards as well as two 2. 5" or 3.5" removable SATA HDDs. The FWA- 6500 also supports four field replaceable unit (FRU) slots that can be used to provide a variety of Ethernet connections. By incorporating packet processing software such as 6WIND’s 6WINDGate*, the system can provide up to 65  Mpps for IP forwarding per socket, depending on the specific use case (Figure 2).  For more details, see the latest Embedded Innovator magazine.



Figure 2.  IP forwarding performance for a 2.4 GHZ Intel® Xeon® processor E5645 running 6WINDGate.  The performance varies based on the number of fast path protocols running in the system.


Deep packet inspection (DPI) is another hot market, one that Emerson Network Power is addressing with its ATCA-7365 AdvancedTCA* (ATCA) blade. Offering the option of two 2.13 GHz Intel® Xeon® processors L5638 processors for use in high temperature telecom environments or two 2.4 GHz Intel® Xeon® processors E5645 for use in cooler environments, the ATCA-7365 provides each CPU with six memory slots for a total of up to 96GB of balanced memory. The blade supports a range of connectivity options, including both 1Gbps and 10Gbps Ethernet. Performance testing, using 6WINDGate software (Figure 3), indicates that for a basic fast path configuration including VLAN, IP forwarding, GTP-U tunneling, flow accounting, and QoS conditioning, an Intel Xeon® processor E5645 can process 2.5 million packets per second (Mpps) per core with an average packet size of 512 bytes (equals roughly 10Gbps of packet ingress and 10Gbps of packet egress per core). The tests indicate that the ATCA-7365 could handle basic communications on 4 x 10G terminations using only 4 of the 12 cores available, leaving ample headroom for additional packet analysis and inspection. When deployed in Emerson’s carrier grade Centellis 2000 2-slot ATCA platform, a system could offer up to 80Gbps of throughput in a 3U form factor. The same core architecture could scale up to 480Gbps in a 3U form factor. (More details on this platform will be available in the September Embedded Innovator newsletter.)


6wind figure 2.png

Figure 3. 6WINDGate* DPI platform for Intel® Xeon® processor-based PCEF.


With its AT8050 single-socket processor blade equipped with an Intel® Xeon® 5500 or 5600 series processor, Kontron is helping TEMs migrate to 40G without a forklift upgrade. TEMs can swap out their existing ATCA node blades and replace them with the AT8050. With up to six cores of processing power, this blade do the work of a dual-socket design, while saving you from having to upgrade a chassis platform to handle the cooling needs of dual-socket replacement blade.  What’s more, the single-socket design frees up an AdvancedMC (AMC) slot for further feature extensions (Figure 4). Kontron provides additional value by integrating the different software components such as OS, BIOS and IPMI, and works with partners like Wind River to pre-validate ‘out-of-the-box’ software kits for faster application development.



Figure 4. The AT8050 offers up to 6 cores and an AMC slot.


Real-time IP media processing typically runs on specialized digital signal processing (DSP) hardware. RadiSys saw an opportunity to deliver carrier-class DSP using software instead, even under heavy load. The RadiSys* Convedia* Software Media Server is a cost-effective and reliable IP media processing platform that can scale to thousands of ports on a single 1U rackmount server platform, or tens of thousands of ports on a fault-resilient bladed chassis architecture. Using commodity server hardware, this Linux-based Session Initiation Protocol (SIP) media server can consolidate the functions of announcement and recording servers, audio and video conference bridges, interactive voice response units (IVR/VRU), messaging equipment, and speech platforms. A special “co-residency” capability allows the integration of a VoIP telecommunications application with the Convedia Software Media Server on a single platform. This feature enables original equipment manufacturers (OEMs) to offer integrated all-in-one telecommunication products at a serious cost advantage. A single Convedia Software Media Server deployment can economically support a broad range of enterprise VoIP applications, including IP PBX, IP Contact Centers, VoiceXML-based IVR, Unified Messaging, or voice/video enterprise-wide conferencing.


To learn more about applications that can benefit from workload consolidation, solutions available from the Intel® Embedded Alliance, and the performance of these solutions, see the white paper Consolidating Communications and Networking Workloads onto one Architecture.


workload_consolidation.pngFor more on building flexible networking solutions, see


Advantech, Emerson, Kontron and RadiSys are Premier members of the Intel® Embedded Alliance. Wind River is an Associate member of the Alliance. 6WIND is an Affiliate member of the Alliance.



Kenton Williston

Roving Reporter (Intel Contractor), Intel® Embedded Alliance

Editor-In-Chief, Embedded Innovator magazine

Follow me on twitter at!/kentonwilliston