Carrier-grade implies software suitable for a carrier, clearly meaning of higher quality in some regard, perhaps in relation to a perceived lower-bar for enterprise software. Does carrier-grade therefore have a higher bar of quality than banking-grade or airline-grade? Google processes over 40,000 internet searches per second, are they carrier-grade?
The truth is that despite some of the examples described in part 1 blog of this blog, public cloud vendors have significant expertise and leverage enormous economies of scale when it comes to running data centers. They are every bit as qualified as telco’s are in providing “carrier-grade” infrastructures. AWS reported over $4.5B in revenue for Q3 2017, with around 30% market share — Microsoft and Google following up in 2nd and 3rd positions, respectively. Gartner estimated in 2016 that Google ran 2.5M servers across all its data centers and products. Those servers aren’t solely dedicated to the Google Cloud Platform (GCP), but the Google expertise, infrastructure and processes are broadly shared, and no one would dispute Google is highly efficient in running data centers.
Given all this, it seems that deploying telco BSS in the cloud could be a highly desirable scenario, it just depends on what objectives you are trying to achieve. You can benefit from proven operating models, an opex vs capex cash flow improvement, and a significant enhancement in agility. However, the total cost of ownership and strategic analysis is unique for every case. It needs to include all the ancillary costs beyond the standard bill of materials, such as: power usage; physical security; real-estate; ancillary equipment such as cooling or cabling; infrastructure operational man power costs; and even third-party software licensing. When looking at the business model this way, it seems clear it is a simpler decision if no existing sunken costs need to be considered: e.g., it is hard to factor in real-estate costs to a TCO analysis if there is an empty data center sitting unused already on the service provider’s books.
The decision as to the right deployment for each Digital Service Provider, cloud or ground, is really based on the unique set of circumstances for that customer. What are the ambitions regarding the infrastructure strategy and future direction? What existing data centers and investments are relevant? Is the customer considering a BSS migration project or starting a new “digital-first” flanker brand to target new market segments? The ideal is that every customer can make a deployment choice given clear options and capabilities that support both scenarios.
This is why at end of 2017 we announced MATRIXX’s availability on the Google Cloud Platform. We chose Google because of their disruptive approach in terms of price, innovation and performance. Capabilities like live migration provide unique operational advantages for certain classes of services, further lowering the operational overhead compared with other public clouds. The disruptive nature of GCP combined with the performance advantage MATRIXX provide a potent Infrastructure and software mix.
Certain service providers may be of a size and momentum that they want to run their own best-in-class data centers and infrastructure (this is certainly one winning strategy), and it further displays the telecommunication industry’s fantastic ability to build and deploy high quality services. Of course, still using MATRIXX for the heart of the BSS!
Like Capital One, some service providers or MVNOs are more focused on the products and consumer services they offer their customers, and might prefer to leverage the cost and efficiency advantages the cloud can offer. For that type of customer, we provide deployment options on Google Cloud Platform.
It’s good to have choice. What is certain, MATRIXX Software offers carrier-grade capability wherever it is deployed.