NFV, Supercharged – Get Inspired & Learn How with Mellanox at MWC16

 
Cloud Computing, Events, , , , , , ,

According to a Light Reading survey, SDN/NFV was ahead of 5G and Internet of Things (IoT) and gained the honor of being the hottest topic at the 2015 Mobile World Congress in Barcelona. Over the course of 2015, we are witnessing more and more NFV projects getting out of the PoC lab environment and moving on to production. In spite of the healthy development, there are still a lot of performance and scaling issues to be resolved. When the rubber hits the road, service providers are realizing that they are not willing to sacrifice on performance and reliability while seeking to get the benefits of virtualization and software defined infrastructure.

 

So where does that leave us?

 

NFV has great promise, but needs to be supercharged to get a performance boost. As the name NFV indicates, the applications here are network functions and the majority of them process packets. Enhancing throughput and packet performance on high-volume servers without adding significant burden to the CPU is the ultimate goal. Achieving this will not only provide a better, more deterministic user experience, but also improve VNF deployment efficiency. This is exactly how Mellanox adds value to NFV and the key theme of our demos at MWC16.

 

Chloe Ma 022216 Fig1

 

So what should you watch out for?

  1. ASAP2

Chloe Ma 022216 Fig2.

 

Accelerated Switching And Packet Processing. This new groundbreaking feature enables the majority of data plane switching and packet processing to be offloaded to the Mellanox NIC, while maintaining control plane operations of virtual switches (first with OVS). Read my contributed article on Light Reading to learn more about the design principle behind ASAP2. Let us show you how you can deliver the highest and most deterministic packet performance, with minimal CPU overhead, while maintaining network programmability.

 

  1. Affirmed Networks MCC over Mellanox 40G interconnect

Affirmed Networks is a leading telecommunications technology supplier with revolutionary Network Function Virtualization (NFV) solutions for the Evolved Packet Core (EPC), vCPE, Gi LAN and SFC Controller. Affirmed’s virtualized Mobile Content Cloud (MCC) software has been designed to run on virtualized high-volume servers.  However when server I/O capacity becomes constrained, application performance may suffer, resulting in under-utilized CPU resources, and excessive server footprint. Mellanox’s high-speed server interconnect solution enhances utilization of infrastructure resources for Affirmed’s broad based virtualized product offerings, enabling optimal application performance as well as space and energy footprint. Also, check out the video in which Ron Parker, Senior Director of System Architecture, discusses how Mellanox helps Affirmed enhance its virtual EPC application efficiency through 40G interconnect.

 

  1. HPE ProLiant Server with Mellanox ConnectX-4 Lx 25G

When it comes to Ethernet speeds you need to know that 25 is the new 10! As a key HPE OpenNFV ecosystem partner, Mellanox has been working closely with HPE to enable 25G as a new option for ProLiant line of servers, and this combination of Mellanox ConnectX-4 Lx 25G NIC on HPE ProLiant server has demonstrated a record-breaking packet performance of 33 million pps with zero packet loss (theoretical maximum packet throughput on a 25G link is 37.2 million pps with 64-byte packets). Best of all 25GbE can use the same SFP cabling infrastructure that is already in your data center. So deploy with 25G to future proof your network and get 2.5X the performance today.

 

Beyond these demos inside the Mellanox booth, there are also other demos at our partners’ booth or venue that have incorporated Mellanox high-speed interconnect as part of their design. Among these, two are particularly worth noting:

  1. Enhanced NFV Scale and Performance with ContexSwitch using Mellanox 40G Interconnect

Chloe Ma 022216 Fig3

 

 

ContexStream, now part of HP Enterprise, is showcasing ContexSwitch at the Israel Mobile Association (IMA) booth stand 2D60 and 2E60 in Hall 2. As a key component of HPE Carrier SDN Solution. ContexSwitch  is a Carrier Grade software OpenFlow switch capable of 160Gb/s throughput with 200M Flows/Rules and 100K Rules/Sec programming. In this demo, ContexSwitch runs on ProLiant DL380 servers equipped with 4X40G redundant connections through Mellanox ConnectX-3 Pro NICs. This system platform provides 160Gb/s I/O throughput perfectly matched to the software capability plus high availability required for carriers to power Service Function Chaining and virtual EPC User plane type of applications.  ContexSwitch uses advanced features such as SR-IOV and Mellanox DPDK PMD to enhance forwarding efficiency and achieve the performance desired.

 

  1. Scalable Processor Cluster (SPC) based on Mellanox Multi-Host Technology

Chloe Ma 022216 Fig4

 

AMD is showcasing its Scalable Processor Cluster (SPC) solution that is based on Mellanox Multi-host technology. Multi-host technology enables designing and building new scale-out heterogeneous compute and storage racks with direct connectivity between compute elements, storage elements and the network, better power and performance management, while achieving maximum data processing and data transfer at minimum capital and operational expenses. Built upon Mellanox Multi-host technology, the AMD SPC provides a significantly more modular, versatile and cost-effective alternative to alternative offerings while still providing high CPU core density. Inter-CPU communication is handled by Mellanox ConnectX-4 NIC, and the demo shows ~30Gb/s aggregated throughput over PCIe interface.

 

 

If you are attending Mobile World Congress in Barcelona this week, be sure to drop by Mellanox booth 5K29 in Hall 5. You will leave feeling empowered with new ideas to supercharge your NFV deployment.

 

 

Comments are closed.