Disrupting the traditional OEM model
A recurring theme in CSP discussions about NFV and SDN is their frustration with traditional OEMs for not moving fast enough to deliver native VNF solutions that run on COTS hardware in a standard way. While CSPs have always been focused on driving suppliers to lower price, the transition to virtualization and software-based networking puts unprecedented pressure on vendors by forcing them to disrupt their own legacy business models. The world’s largest telecom equipment vendors – Ericsson, Huawei, and Nokia – will one day likely be compelling case studies that illustrate how companies can either embrace disruption to reinvent themselves or succumb to disruptive forces.
The traditional OEM business model is based on selling dedicated hardware appliances and licenses for the software that runs on them. This transaction is too restrictive for today’s networking needs because the hardware provides an inflexible, minimum amount of capacity that can only be expanded by purchasing additional equipment. Operators also become locked into supplier relationships due to the need for ongoing hardware maintenance contracts. And the entire process of procuring, designing, integrating and deploying network equipment is too slow and stifles innovation.
With the adoption of standard open interfaces, NFV and SDN change the traditional supplier business model from hardware contracts to software licenses. Rather than purchasing dedicated appliances for a specific amount of capacity, mobile operators will buy software licenses and pay for what they use in the network. Traditional OEMs are struggling to adapt to new software licensing models. Pure software vendors that are not burdened by legacy hardware businesses are well positioned to deliver the native virtualization solutions that mobile operators require. But it is not easy for operators to extricate themselves from long-term supplier relationships. Since mobile operators are not going to rip out entire networks overnight and replace them with virtualized implementations, they will depend on their existing suppliers of legacy equipment during the transition.
But the financial incentives for moving just one network element to the software are overwhelming. In equipment purchase prices, typically 10% is for maintenance services. CSPs can implement the same function in an overlay virtualized network for less than what they are paying in maintenance costs for the equivalent network appliance. The implementation of a VNF in the mobile network would pay for itself within one year. Whether it’s an instantiation of vBBUs in a Virtualized RAN deployment or implementing vEPC, mobile operators will reap the cost savings benefits of NFV one VNF at a time.
In the case of a vBBU deployment, a small operator with 10,000 cell sites would pay in the realm of $720 million for BBU functionality, of which $72 million would be maintenance fees. If the same operator deployed vBBU (assuming 240,000 cores at 100% utilization), the price for the Intel machines and software licenses would be$46 million – significantly less than the cost of a BBU appliance contract and even less than the maintenance fees.
Without the constraints of legacy business models, new software vendors are poised to disrupt the supplier landscape and take market share from traditional OEMs. With software assets that are natively designed for cloud environments, the new breed of software supplier can deliver the innovation, agility and costs savings that mobile operators need to survive.