Hyperscaler is recently added, new term in Telecom World.
In computing, hyperscale is the ability of an architecture to scale appropriately as increased demand is added to the system. In short, when we say hyperscalers, we mean huge companies like Google, Microsoft, Facebook, Alibaba and Amazon who owns huge datacenters and provide scalability to host growing data subscribers and traffic.
In recent years, we are seeing growing partnership between Hyperscalers and Telco Providers in edge computing as shown at https://stlpartners.com/research/telco-edge-computing-how-to-partner-with-hyperscalers/) (See figure below )
Relationship between Hyper-scaler and Telco company is edge computing is based on following principles
- Sharing customers for use-case such as MEC and Private VoLTE
- Sharing Datacenters for Edge computing use-cases.
- Developing Edge applications by using each other’s software development ecosystem
Apart, from collaboration in Edge computing, there are other areas where Telcos can work with Hyperscalers e.g. migration of networking and non-networking applications to Hyperscaler cloud for cost and operational efficiency. By definition Networking applications are Applications which serves directly or in-directly basic telecom & data functions such as gateway routers (Packet-GW, UPF), control plane elements(MME), authentication servers (HSS) etc. Non-Networking applications are Applications which serves business functions such as billing & mediation services, human resource management applications (SAP), etc.
Key consideration for Hyperscaler migration for Telco
Following the key considerations for Telco's Hyperscaler migration:
- Transitioning from Cost to Expense financial model
- Adaptation of Hyperscaler Billing & subscription based hierarchy
- Controlling network egress bandwidth Cost
- Defining CI/CD pipelines for external and internal application using Hyperscaler devops tools
- Controlling cost of High availability and Disaster recovery in Hyperscaler enviroment
(to be continued)
