5G pricing - How would Service Providers monetize 5G investments?

5G deployments and commercial launches have started in many parts of the world. Service providers are spending several billions of dollars to build the 5G network. But, most service providers are still in the "learning phase", trying to understand how to monetize their 5G investments. "How is 5G going to be priced?" - is still a million dollar question for many in the industry. In fact, 5G pricing is still a puzzle for even the service providers. This article explores how 5G can be priced by the service providers.

A quick look at the history

"Study the past, if you would define the future", said Confucius, a great philosopher. So, it is important to look at how Service Providers priced their previous generation wireless technologies, before looking at 5G pricing.

Traditionally, service providers charged their wireless customers based on the amount of bandwidth they consumed. For example, a customer using 2Gb of data a day is priced higher than the customer who is using 400 Mb of data, a day. But, that pricing model isn't going to be relevant for 5G, as 5G caters to a variety of use cases such as IOT and Autonomous Cars. For example, in an IOT deployment, several 100s and 1000s of sensors / devices are going to push small chunks of data at periodic intervals to the network or to the cloud. The aggregate volume of data pushed to the network/cloud isn't going to be as high as 2Gb a day. However, the network has to support several 100s and 1000s of active sensors/devices at a given time. So, service providers cannot make a lot of money if they charge customers just based on the bandwidth consumed. They'll have to be more creative at charging the customer, month-on-month basis.

How Service Providers are pricing 5G today?

US has been leading the 5G wave. So, let us look into how the major service providers in US are pricing their 5G services.

Verizon charges $70 per month for new customers signing up for Verizon 5G Home services. They are charging $50 per month for existing wireless customers who sign up for Verizon 5G Home services. The good news is, there is no data cap/limit for 5G Home customers. So, they can do unlimited data downloads/uploads.

For 5G mobile services, Verizon has mentioned that it is going to charge an additional $10 per month, when compared to the 4G customer.

However, even the best Verizon's Unlimited plan allows a customer to access up to 75Gb of data a month, beyond which it may lower the speed of data access.

AT&T's initial launch of 5G services comes with a $70 per month plan and with a 15GB data cap/limit.

T-Mobile has mentioned that they won't charge extra for a 5G connection. T-Mobile is planning to generate additional revenue through 5G by offering home broadband service and Internet of Things (IoT) capabilities.

How can Service Providers price their 5G services?

There are a few indications from leading service providers on how they are planning to price their 5G services. For example, Hans Vestberg in his keynote address at the Consumer Electronics Show (CES) talked about the 8 currencies of 5G. He talked about 8 key parameters based on which the 5G services can be priced such as Peak Data Rate, Mobile data volume, Rate of Mobility of the device/user, number of Connected Devices, Service Deployment time, Service Reliability, Energy Efficiency and E2E Latency.

In the recent earnings call, AT&T CEO Randall Stephenson mentioned that they are going to charge 5G customers based on the speed of network connection. For example, customers who are availing a 1 Gbps 5G service would be charged higher than customers who are availing just 50 Mbps service.

1. Pricing based on Bandwidth consumed - This is the traditional pricing strategy and some service providers would continue to provide 5G wireless plans based on customer's bandwidth consumption. Once the customer reaches their maximum upload/download limits, the service provider would switch the customer to a slower tier of service.

2. Pricing based on network latency - There are services such as Autonomous Cars and Robotic Surgeries that are latency sensitive. Customers and Enterprises using such services would avail plans that support different latency tiers. For example, a customer signing up for 1 millisecond latency would be charged higher than the customer who signs up for 5 milliseconds or more latencies.

3. Pricing based on number of devices  - In an IOT deployment, pricing of the 5G services is going to be based on the number of devices that connect to the network. For example, a farm land would have several 10s or 100s of sensors / devices actively monitoring parameters like weather, water levels, soil quality, pesticide levels etc., and pushing such data to the cloud. In such deployments, access to 5G services will be charged based on the number of devices that connect to the 5G network.

4. Pricing based on network speed - Services such as online gaming, Augmented Reality / Virtual Reality (AR/VR) would require a network that performs well and can support higher data speeds. Customers who use AR/VR or online gaming services can sign up for a pricing plan that offers higher network speeds.

5. Pricing based on Network Slice  - Network slicing is a new technique introduced in 5G, to partition the physical network into multiple virtual network slices, based on parameters such as speed, latency and security. An Enterprise customer can avail a dedicated network slice that offers a secure SD-WAN connectivity between its branch offices. This would improve the quality of the SD-WAN service for the customer..

6. Pricing based on tenant space in the MEC infrastructure - Multi-access Edge Cloud (MEC) infrastructure opens up the service provider edge for hosting a variety of MEC applications such as video streaming, video surveillance, IOT and AR/VR applications. Service providers can charge their partners and customers for the MEC tenant space. The various MEC use-cases would enable service providers to create additional revenue streams.

7. Free 5G service (subsidised by OTT players) - Considering the growth of OTT (Over-The-Top) services, online service providers are even ready to subsidize the connectivity to Internet for their customers. Google today is rolling out free WiFi services in railway stations, in countries like India. In the future, online service providers like Google may partner with telecom service providers to offer free 5G services to their customers. 

8. Pay using Bitcoins - Imagine a day where you'll have to play online games and earn bitcoins that you can use for paying your monthly 5G services bill. Not sure whether we'll ever get to this. However, considering the amount of time subscribers spend in online gaming applications (and in earning bitcoins), this could also be a reality.

In the next three years, we would have more clarity on how service providers are going to monetize their 5G investments. In the mean time, we can ponder over the various 5G pricing options available and service providers can start experimenting with a few.

List of OpenFlow Controllers for SDN

OpenFlow SDN Controller is a critical component in the SDN architecture. Though OpenFlow protocol has lost its shine in the last few years, there are still a number of SDN controllers that support OpenFlow. This article captures the list of open source and commercial SDN controllers that support OpenFlow protocol. If you are interested to know the OpenFlow controllers in the market for SDN deployments, you can find them below:

(Also read: Is SDN really dead?)

Also read: 

Software Defined Networking Basics / 101
SDN Architecture - Protocols & Components
OpenFlow Test Suites & Tools

OpenDaylight is THE most popular SDN controller in the market, which supports OpenFlow protocol. It supports at least a dozen southbound APIs including protocols such as OVSDB, NetConf and PCEP to configure / manage network devices. There are multiple vendors such as Lumina Networks, Ericsson and Cisco selling OpenDaylight based SDN controller solutions. You can learn more about OpenDaylight at https://www.opendaylight.org/.

NOX (Network Operating Systems) is the original OpenFlow controller, and facilitates development of fast C++ controllers on Linux. It is an open source implementation and can be downloaded from http://www.noxrepo.org/

POX is a Python based open source implementation for on Windows, Mac OS, or Linux. It is used primarily for research and education and can be downloaded from http://www.noxrepo.org/

Beacon is a fast, cross-platform, modular, Java-based OpenFlow controller. Beacon is an open source implementation and claims to have been used in several research projects, networking classes, and trial deployments. Beacon seems to be currently powering a 100-vswitch, 20-physical switch experimental data center and running for months without downtime. It can be downloaded from https://openflow.stanford.edu/display/Beacon/Home

Big Network Controller is the network application platform for the Open Software Defined Network (SDN), providing unified network intelligence, enterprise-class scalability and high availability, and a platform to deploy a wide range of network applications, including data center network virtualization. Big Network Controller uses industry standard protocols, like OpenFlow, to create a common abstraction and universal data model for the underlying network data plane elements.  Visit http://www.bigswitch.com/ for more information. (Product EOLd)

Maestro is a scalable control platform written in Java which supports OpenFlow switches. It is distributed as an open source software. Maestro provides interfaces for implementing modular network control applications. Maestro is a platform for achieving automatic and programmatic network control functions. It can be downloaded from http://zhengcai.github.io/maestro-platform/

Floodlight controller is an enterprise-class, Apache-licensed, Java-based OpenFlow Controller. Floodlight supports a broad range of virtual and physical OpenFlow switches. It can manage multiple islands of OpenFlow switches, a common deployment scenario. It forms the core of Big Switch's OpenFlow Controller. Yes! Big Switch open sourced Floodlight. It can be downloaded from http://floodlight.openflowhub.org (No activity for a long time in the community)

Ryu is an Operating System for Software Defined Networks. Ryu aims to provide a logically centralized control and well defined API that make it easy for operators to create new network management and control applications. It is developed and maintained by NTT labs. It can be downloaded from http://osrg.github.com/ryu/

Faucet is a compact open source OpenFlow controller, which enables network operators to run their networks the same way they do server clusters. Faucet moves network control functions (like routing protocols, neighbor discovery, and switching algorithms) to vendor independent server-based software, where those functions are easy to manage, test, and extend with modern systems management best practices and tools. Faucet controls OpenFlow 1.3 protocol. https://faucet.nz/ 

The HP Virtual Application Networks SDN Controller is the center-piece of the SDN architecture. The controller offers a dynamic control plane with the intelligence to automate and program the network to enable network agility. It has full support for the OpenFlow protocol and  has Open APIs to enable third party SDN Application Development. More details at http://h17007.www1.hp.com/us/en/solutions/technology/van/index.aspx

NEC ProgrammableFlow Controller is designed to work with low cost Open Compute Project (OCP) inspired network equipment. It supports a packet processing pipeline capability to ensure an efficient, high performance and flexible network architecture. It supports up to network 10,000 switches. More details at https://www.necam.com/sdn/Software/SDNController/

Lumina SDN Controller is one of the popular OpenDaylight based SDN controllers (spun-off by Brocade). It creates a common control plane over multiple domains so that services can be deployed using a single set of applications. The same service can be instantiated over a routed domain, a switched domain and multiple vendors simultaneously.  Network Convergence is a key differentiator for the Lumina SDN Controller. More about Lumina SDN Controller at https://www.luminanetworks.com 

The IBM Unified Controller provides an OpenFlow-based network fabric with centralized control of network flows and unlimited virtual machine (VM) mobility—implemented in enterprise-class software. The controller software automatically and continuously discovers the OpenFlow network topology and maps physical and virtual traffic flows across any OpenFlow-based data center network environment. IBM isn't still an established player in the industry for SDN controllers. More details at https://www-01.ibm.com/support/docview.wss?uid=isg3T7000691&aid=1

Ericsson Cloud SDN Controller provides seamless intra & inter-datacenter connectivity for virtual, physical and container based workloads. It uniquely combines an industrialized OpenDaylight controller with advanced routing capabilities. It is delivered as part of our system-verified NFVI (NFV Infrastructure) solution and provides network automation services to NFV workloads. More details at https://www.ericsson.com/en/portfolio/digital-services/cloud-infrastructure/cloud-sdn

Cisco Open SDN Controller is a commercial distribution of OpenDaylight that delivers business agility through automation of standards-based network infrastructure. It abstracts away the complexity of managing heterogeneous network environments to improve service delivery and reduce operating costs. More details at https://www.cisco.com/c/en/us/products/cloud-systems-management/open-sdn-controller/index.html

Also read: 

Software Defined Networking Basics / 101
10 SDN Startups
SDN Architecture - Protocols & Components

Applications & Use Cases for MEC / Edge Cloud

Multi-access Edge Computing (MEC) is one of the several pillars that is holding the 5G monument, together. MEC is now synonymous with Edge Computing and Edge Cloud. Edge Computing infrastructure is not a new concept to Service Providers. Even in the past, Service Providers did have some foot print of compute infrastructure at the edge, providing services such as caching, video processing and WAN optimization. However, that infrastructure was a black box to many - allowing only a handful of partners to deploy their servers and run their applications. With the introduction of 5G, MEC opens up that infrastructure for any thirdparty to run their applications, closer to the edge and closer to the users. MEC makes that infrastructure more dynamic and programmable by exposing APIs. MEC helps 5G to keep up the promise of latency, which is one of the 8 currencies of 5G. MEC supports a wide variety of applications and use-cases - some are specific to 5G usecases and many of them are generic use-cases. In this article, we'll look at the top MEC applications / use-cases, that are getting traction in the industry.  (Also read: Differences between Cloud, Fog and MEC - Removing the Mist)

Video Streaming & Entertainment Services from the Edge - Video streaming is going to be one of the biggest use cases supported by MEC. According to Cisco's Visual Networking Index 2019 report, IP video traffic will be 82 percent of all IP traffic (both business and consumer) by 2022, up from 75 percent in 2017.  In addition, Internet Video will grow by 33% year-over-year. For example, in a stadium, audience can record and share high quality video streams from the event to their friends in social media. Passengers traveling by flight or train will be watching HD video shows or movies from their mobile devices. When an MEC infrastructure is used for hosting applications that accelerate video steaming and entertainment services, it reduces the latency for users and improves user experience.

Video analytics applications - With more and more people watching online video, mobile devices are continuously generating lots and lots of data about user's viewing patterns, user's geographic location, user's interests etc., Today, such video analytics data get transferred from the user equipment (such as laptops, mobile phones, Set Top Boxes) to a centralized cloud infrastructure, for processing to derive meaningful insights. This increases the overhead on the network links and also increases the latency of services. For example, based on video analytics data, a video streaming solutions provider can locally process the data and customize the playlist based on the interests of the local community. By hosting video analytics applications at the MEC, such processing of large volumes of data can be done at the edge, versus processing them at a centralized cloud infrastructure. This saves a lot of time as well as costs for the streaming service provider.

AR/VR Applications - Enteprises such as an airport could use MEC for hosting applications that serve advertisements and provide location based and augmented reality services.  For example, a AR enabled mobile application can guide a user to the boarding gate within a terminal or to a dining area, by showing the path/directions using the live map. These applications can be locally hosted and managed by the airport services provider. An industrial plant could use MEC for hosting applications that are used for training the workforce using Augmented or Virtual Reality. 

Surveillance applications - Today, it's tougher to spot a water dispenser, but easier to spot a surveillance camera. Many enterprises and communities are installing surveillance cameras in several locations to improve their monitoring ability and to reduce crime. A crowded street fitted with multiple high resolution surveillance cameras are continuously capturing and streaming live video feeds to the centralized cloud. If the security team wants to upload a team of known criminals into the system and find their whereabouts, they can use the MEC infrastructure to host such security services. Instead of having to process the live feed at a centralized location, it can be locally captured/processed at the MEC infrastructure. This would help the security personnel to respond to any security needs in the community, in a timely fashion.

IOT Edge Gateway - Internet of Things (IOT) is enabling a new set of use cases for MEC. IOT Edge Gateway is one such function in an IOT architecture, which typically hosts applications that gather lots of data coming from the various sensors and devices in the IOT network. Such data gathering applications can be hosted on MEC. In addition, applications that process the data to generate IOT analytics and insights can also be hosted in MEC.

Network Probes - Today, network probes are initiated from several 100s of edge locations to monitor / measure the performance of a network. These services are hosted in special purpose built applications. Instead, such network monitoring and probing applications can be run on 1000s of MEC locations. This helps service providers to monitor not only the availability and health of the network, but also the quality of service offered to the various applications and services.

WLAN Controller - In an Enterprise deployment, MEC can help in hosting enterprise network management applications such as a WLAN controller application. Today, WLAN Controller is a dedicated appliance within the Enterprise network. MEC can help Enterprises to migrate that functionality to the edge cloud instance. MEC infrastructure in an Enterprise can also be used for hosting policy management services. 

Enabling Autonomous Cars - Autonomous cars would need real time information about traffic congestion, weather, road blocks, traffic diversions etc., Such information can be served from applications that are running on the Edge Cloud in the MEC infrastructure. An autonomous car has only a fraction of a millisecond to think and make realtime decisions. Applications hosted in MEC infrastructure can help autonomous cars to have real time access to important data. 

Edge Caching - Edge caching is one of the traditional services that was hosted and run in a service provider network. Mostly, such services were hosted in purpose built appliances supplied by vendors such as Google and Facebook to improve their application performance. Also, CDN service providers such as Akamai install their appliances in the service provider network to reduce the latency of services delivered by the CDN. Now, with the advent of MEC, such caching services can be hosted in the local MEC infrastructure to improve the performance of web/mobile applications, video/music streaming applications and services such as social media, online gaming and AR/VR. (Also read: Components of a Content Delivery Network (CDN))

Traffic Optimization - MEC can host applications that are used for traffic optimization. For example, services such as WAN acceleration, video compression, video transcoding can be done at the MEC edge. This would reduce the volume of data that gets transferred over the network, and hence improves the speed of services.

Security Applications - There are different types of security applications that can be hosted in a MEC infrastructure. For example, in an IOT environment, MEC can host IOT security applications, that provide a layer of separation between the IOT devices/sensors/controllers and the cloud. In an Enterprise environment, MEC can host Enterprise security and policy management applications. For example, traditionally Enterprises did not have any control over the traffic going over the mobile wireless network (especially, on the mobile devices issued by the company to the employees). With 5G's service based architecture and the support for APIs, Enterprises can get access and control over the traffic going over the mobile wireless network. That enables the Enterprises to have consistent policies between the mobile wireless network and the local WLAN infrastructure. (Also read: Will WiFi get killed by 5G?)

SDN Switches - Black, White and the Gray

Marcus Luttrell, a former US Navy SEAL once said "War is not black and white; it's gray. If you don't fight in the gray area, you're going to lose". Well, the hardware equipment vendors have realized, that in the era of SDN, if they don't fight in the gray area, they are going to lose. Several small startups and open source communities waged the SDN war on traditional equipment vendors, by fighting their proprietary black box switches with the white box switch. The equipment vendors retaliated back, by introducing gray box switches. 

(Also read: SDN white box switch vendors in the market)

If you are confused about the differences between the black and the white and the gray box switches, this article will rescue you. Equipment vendors are inventing new terminologies to keep them afloat - so that, they can continue to sell hardware equipment to enterprises, data centers and service providers. You can't blame them, as hardware is where they make money. 

Now, let us cut to the chase.

Black Box Switches

Black Box switches are the legacy network switches. Both the hardware and software running on those switches were developed by the same vendor. Black Box switches are typically propriety network equipment, using proprietary hardware and proprietary software. Customers using the Black Box switches cannot write their own applications. They've to go to the device vendor to get any customizations - involving long lead time and huge costs. You can compare this to the good old days where HP, IBM and Sun sold their Unix operating systems bundled with their proprietary hardware. Today, almost every network equipment vendor claims support for open APIs and SDN protocols in their black box switches. Black box switches are costly and customers get locked-into proprietary hardware and software. 

White Box Switches

One of the goals of SDN is to challenge the proprietary Black Box switches in the market. Thus came the White Box switches. White Box switches are typically SDN enabled switches. One of the major differences between the white box and black box switches is the decoupling of the software from the hardware. Customers would have an option of picking the software (network operating system / applications) of their choice and run it on a hardware of their choice. On top of it, White Box switches can be managed using open APIs or SDN protocols such as OpenFlow or OpenConfig. You can compare this to Microsoft Windows operating system running in Dell, HP or IBM hardware. There is no tight coupling between the hardware and software - however, they interoperate with each other. White Box switches are ideal for data center deployments, where reliability is achieved by flexibly moving the workload to a different server, rack or even datacenter. White Box switches are typically low cost and do not meet the service providers' "carrier grade" requirements. 

(Also read: SDN white box switch vendors in the market)

Gray Box Switches

In most cases, the white box switches came from vendors such as Advantech, Accton, Dell and HP. This posed a huge threat to the traditional network equipment vendors such as Juniper and Cisco. So, they started re-incarnating their traditional offerings and branded them as "open solutions". For example, they decoupled their network software from hardware and started licensing them to customers. They also invented a new breed of offering called the gray box switch. The gray box switch allows you to not only run the proprietary vendor software, but also allows you to run other vendors' software. Juniper's NFX250 Network Services Platform is a good example of the gray box. NFX250 is a Universal Customer Premise Equipment (CPE) platform that allows customers to run  other vendors' Virtual Network Functions (VNFs) in addition to supporting Juniper's VNFs. SD-WAN like use-cases require enterprise customers to run multiple VNFs on the edge to support services such as routing, firewall and WAN optimization.  Gray box switches serve as an ideal platform for running such applications. Gray box switches are "carrier grade" and they meet the service provider's reliability requirements. Hence, they are costlier than the whitebox switches.

So, which one of these network equipment is the best for your use? Actually, that depends on your use case, budget and reliability requirements. 

What makes 5G networks faster?

In his keynote speech at Consumer Electronics Show (CES), Verizon's CEO Hans Vestberg talked about the Eight 5G Currencies and mentioned that 5G speeds are expected to touch 10 Gbps, soon. 5G technology supports speeds of 1 Gbps today, which is still twice the speed offered by the best 4G network. With such high speeds, you should be able to download a HD movie in a fraction of a second. What makes 5G faster than the previous generation technologies? How can 5G networks deliver better speed? This article will look into the nuts and bolts that make 5G faster than other technologies. 

There are several technological advancements in 5G architecture, which makes it significantly faster than the previous generation wireless technologies such as 4G and 3G. Some of them are given below:

1. Spectrum - 5G's spectrum plays a significant role in improving the 5G network speed. 5G networks support 3 different ranges of spectrum - low band spectrum which is below 1 GHz, mid band spectrum from 1 GHz to 6 GHz and high band spectrum 24 GHz / 30 GHz to 300 GHz. The high-band spectrum is also known as millimeter-wave. Lower frequencies typically travel long enough and can propagate deeper than higher frequencies. Lower frequencies offer a better coverage. Higher frequencies will travel short distance, however can offer higher bandwidth. 5G has to offer both better coverage and higher speeds - and hence, millimeter wave is very important for 5G to offer higher speeds. (Also read: 5G Technology Overview and FAQ)
2. Massive MIMO - 5G adopts massive MIMO technology where a few tens to several hundreds of antennas are used for propagating radio signals. The Massive MIMO enabled radio devices use algorithms to calculate the best possible transmission route through the air, to reach the receivers without any interference or with reduced interference.  When the interference goes down, efficiency of transmission goes up and hence, Massive MIMO helps 5G to provide higher speeds.
3. Network Slicing - Many folks may wonder, how network slicing can influence 5G's speeds. However, Network Slicing carves out a dedicated capacity for you in the network. It is like someone building an expressway and carving out a dedicated lane for you and blocking others from entering the lane. You don't have to worry about network congestion and network slowness, as the capacity is dedicated completely for you. Hence, Network Slicing influences 5G's data speeds. (Also read: 5G Network Slicing - What, How and Why?)
4. Multi-access Edge Computing (MEC) infrastructure - MEC infrastructure influences both the latency of transmission and the speed of transmission. It is always quicker and faster to buy something from a local shop across the street, than buying something online, where your order gets shipped from a central warehouse. MEC does the same for 5G network. It offers compute infrastructure closer to the edge of the network, where service providers can host their applications or partner applications. Imagine a customer in San Jose, downloading a HD video from a datacenter in New York, vs. having to download it from a local MEC infrastructure in San Jose. Thus, MEC helps the service providers to offer higher speeds for services accessed through 5G network. (Also read: Top MEC vendors to watch)
5. Virtualization - In the traditional network architecture, most of the network functions were served by special purpose hardware equipment. Service providers had separate hardware for functions such as firewall, router, cache and policy management. However, 5G services are cloud native. Most of the core network infrastructure for 5G network, runs in the telco cloud. A virtualized 5G network helps the service providers to deliver higher speeds in more than one ways. First of all, services that are inter-dependant with each other (for ex., caching server and a streaming server) can run in the same rack/hardware in the datacenter. This would reduce the number of network hops between these network functions. Second of all, service providers can dynamically scale their network infrastructure based on the load. Thus, virtualization allows the service providers to offer better 5G speeds. 

We still have a long way to achieve the vision of getting 10 Gbps wireless network speed. However, with several network equipment vendors and service providers aggressively investing their R&D dollars in 5G - you can expect the speeds to go up more, in the coming years. 

SDN Whitebox switch vendors and startups in the market - A Closer look

One of the promises of SDN technology, is to detach the tentacles holding the hardware and software together, and make the network equipment more open for innovation. When the SDN movement started 10 years back, a number of startups sprouted to develop open networking operating systems and white box switches for data centers. They rolled up their sleeves to make switching software more open and more programmable. This article will look at some of the SDN whitebox switch vendors and startups in the market, that have evolved over the last decade.

BigSwitch Networks (https://www.bigswitch.com/)

When was the company founded?

Big Switch Networks was founded in 2010. Big Switch was known for its Floodlight SDN Controller, which got open sourced in the year 2012.

What are the company's products?

Big Cloud Fabric is the key product offering from Big Switch. It enables virtual private cloud (VPC) based logical networking, delivering network automation and visibility to both on-premise enterprise clouds and multiple public clouds. It is offered in two variants - Enterprise Cloud and Public Cloud.

The company also offers other products for monitoring and analytics.

What are the company's differentiators?

Big Cloud Fabric uses virtual private cloud (VPC) as the basic construct in both on-premises enterprise clouds and multiple public clouds. It provides a consistent network management and operations management capabilities, simplifying Day-0 through Day-2 operations for customers.

Who are the key customers?

Big Switch claims to have customers including 20% of the Fortune 100 companies. Some of its key customers are Verizon, VMWare, Visa and T-Mobile.

What is the current status of the company?

Big Switch has raised $111.7M funds and its estimated annual revenue is $36M. Some of its investors include Dell Technologies Capital, Intel Capital and Khosla Ventures.

Cumulus (https://cumulusnetworks.com/)

When was the company founded?

Cumulus was founded in 2010.

What are the company's products?

Cumulus is one of the pioneers in the industry offering open networking operating system for whitebox switches. Cumulus Linux is their key product offering, which is an open network operating system that allows you to automate, customize and scale your datacenter networks.

What are the company's differentiators?

Cumulus provides support for 100+ hardware platforms including industry standard hardware that are Open Compute Platform (OCP) and ONIE (Open Networking Install Environment) compliant. Cumulus is the first white box software vendor to add support for Minipack, Facebook's latest OCP compliant reference design.

Who are the key customers?

Cumulus claims to have 34% of the Fortune 50 companies. Some of its key customers are Verizon, NTT, Yahoo Japan, Qualcomm and Paypal.  

What is the current status of the company?

Cumulus has raised $129M funds and its annual revenue is $15M. Some of its investors include Sequoia Capital and Telstra Ventures. 

Pica8 (https://www.pica8.com/)

When was the company founded?

Pica8 was founded in 2009.

What are the company's products?

Pica8's key product offering is PICOS, an open networking software. PICOS provides tightly coupled control planes, giving network operators non-disruptive control of their enterprise applications; deep and dynamic traffic monitoring; and attack mitigation – all in real time.

What are the company's differentiators?

PICOS supports a functionality called CrossFlow, which is only available from Pica8. CrossFlow functionality tightly couples L2/L3 control plane and classic “SDN” control plane to give network operators real-time, non-disruptive control of their enterprise applications; deep and dynamic traffic monitoring; and even attack mitigation without interrupting flow processing or requiring switch restarts.

Who are the key customers?

Pica8 claims to have over 900 customers in 30+ countries. Some of its key customers include General Electric, Stratedge and Edge Core. What is the current status of the company?

Pica8 raised $18M in funding and its current annual revenue is approximately, $3.2M per annum. Some of its investors are Vantage Point Capital Partners, Pacific Venture Partners

Plexxi (https://www.plexxi.com/)

When was the company founded?

Plexxi was founded in 2010.

What are the company's products?

The company's key product offering is Plexxi Switch, which enables customers to build public and private clouds for the next era of IT. Post the HPE acquisition, this product is getting integrated into the HPE Composable Fabric ecosystem.

What are the company's differentiators?

Plexxi's product supports Composable infrastructure delivering fluid pools of networking resources that can be composed and recomposed as needed, based on the network load.

Who are the key customers?

Some of Plexxi's customers include Arrow Electronics, SafeGuard Scientifics and Jefferies Financial Group. 

What is the current status of the company?

Plexxi got acquired by HPE in May 2018. The company raised a total funding of $83.4M, with an estimated revenue of $12.2M per annum. Some of its investors include Matrix Partners and Lightspeed Venture Partners. 

Pluribus (https://www.pluribusnetworks.com/)

When was the company founded?

Pluribus was founded in 2010.

What are the company's products?

Pluribus' key product offering is Netvisor ONE Operating System, which is a Layer 2/3 Switching Optimized to meet the requirements for Distributed Enterprise and Service Provider Networks.

What are the company's differentiators?

Pluribus supports a peer-to-peer architecture, which eliminates the need for SDN controller and simplifies the overall network architecture. It supports Open Compute Platform (OCP) and Open Network Install Environment (ONIE) standards.

Who are the key customers?

Pluribus claims to have 2 out of the top 10 computing companies, and 1 out of the 5 global banks as its customers. Some of its key customers are Cloudfare, Steelcase and Tibco. What is the current status of the company?

Pluribus has raised $92.5M, so far from investors. It's current revenue is $20M per annum. Some of its investors are Ericsson Ventures, Menlo Ventures, AME Cloud Ventures and CBC Capital.

NoviFlow (https://noviflow.com)

When was the company founded?

Noviflow was founded in 2012.

What are the company's products?

Noviflow's key product offering is NoviWare™ Network Operating System software which can be deployed in network switches, WAN IP/MPLS routers, network appliances and other high-bandwidth forwarding planes. It is compliant with OpenFlow 1.3/1.4 and 1.5 protocol versions.

What are the company's differentiators?

Noviflow offers the industry’s most complete implementation of the OpenFlow 1.3/1.4/1.5 protocol standard.

(Also read: Open Flow Switches & Switch Vendors)

Who are the key customers?

Noviflow's key customers are Telstra group and Chunghwa. They are also partnered with companies like Fortinet and Lumina Networks to offer combined solutions to customers. 

What is the current status of the company?

Noviflow has raised a funding of $9M, so far. They are currently generating a revenue of less than $1M per year. Some of its investors are Fonds de solidarité FTQ and Plug 'n Play. 

(Credits: The source data for this research is from www.owler.com, www.crunchbase.com, www.hgdata.com, www.featuredcustomers.com and the respective company websites)https://tech.ginkos.in/2012/12/openflow-switches-vendors-for-sdn.html