Hyperscale Cloud Vendors Challenge Communication Service Providers: What’s Your Next Move?

Communication Service Providers (CSPs) – whether Internet, carriers, cableco, satellite, or managed service – are playing a critical role in the next era of growth and innovation. They are enabling the next generation of user/customer/digital experiences that are in such great demand.

However, they are now challenged to move beyond their core connectivity business and introduce value added services that are unique and profitable.

If you are a CSP, you know this is a difficult challenge to navigate. What you sometimes view as a major hurdle to innovation – your legacy infrastructure – may in fact be your greatest asset. But before I get into that, let’s look at what’s happening in the industry to provide some context.

Increased Competition

There has never been a time of such increased competition as now. CSPs are not only competing against each other, but now hyperscale cloud vendors are intent on displacing you by leveraging their leading-edge infrastructures and seeming limitless funds to invest.

Hyperscale vendors are prioritizing fast, reliable connectivity. For example, one company is focused on driving down the cost of connectivity with technology that has immense disruptive potential. Instead of building cell towers, the firm will resell access to different mobile carrier networks — enabling consumers’ phones to use the best signal available at any moment. That could transform the economics of wireless service – and carriers could become “invisible wholesalers competing to offer access at the cheapest rates.”[i]

What’s even more challenging is that it’s hardly a level playing field. Hyperscale is largely unregulated, while carriers are heavily regulated and overseen by the government. While hyperscale vendors are buying equipment for enormous datacenters packed with hundreds of thousands of servers, carriers have localized real estate with infrastructure from a few servers in half racks to dozens of racks.

Finally, hyperscale companies are buying at such large volume that they can influence cost points that others cannot. Carriers aren’t buying at the same volume so it’s extremely difficult to achieve the same scale, cost controls or operational efficiencies.

Making the Next Move

So what moves are now available to you? We know there are two fundamental (and often competing) areas at play here.

  1. Efficiency— How to drive down costs, increase profits, and heighten security while enhancing the customer’s perceived experience.
  2. Innovation— How to address new market conditions, changing customer preferences, and digital disruption with new and compelling offers.

Despite some of the advantages hyperscale organizations possess, CSPs do possess two advantages that others lack.

One is a differentiated understanding of ‘the edge.’ Carriers have been delivering capability at the edge for decades. Between fiber to the home and 4G wireless connectivity, you understand how to successfully deliver services to individual consumers. You’re also skilled in leveraging local infrastructure to enhance service delivery for a home or neighborhood.

You probably also have real estate in your favor. There are millions of point-of-presence locations on towers, streets and office buildings that provide edge capacity for emerging workloads such as self-driving cars, artificial intelligence, virtual reality and the Internet of Things.

Hyperscale vendors, by contrast, struggle with these factors because they typically have centralized, massive data centers in a handful of locations. But these cloud companies recognize that carriers have this advantage, and they are eager to acquire local capacity. This will take time, and they must make decisions around buying, building, or partnering for capacity.

CSPs, this is your opportunity to strike. Moving with speed, nimbleness and adaptability, you can broaden the definition of cloud to encompass:

  • central compute / storage / networking
  • transport links
  • edge compute / storage / networking
  • communications infrastructure, including fiber and radio

How Dell EMC Can Help

While this may all sound optimistic, we know it isn’t an easy task. Your CTO office may be asking to move to 20 kilowatt racks like the competition; your facility team may be pushing back with power limitations in certain locations; and your corporate real estate folks (who often make the final decisions) may not be motivated in the same manner as you, especially if they’re dealing with regulatory compliance. As a result, your ability to adapt may suffer.

This is where Dell EMC’s Extreme Scale Infrastructure (ESI) division can help. We have been producing advanced infrastructure improvements to some of the world’s largest hyperscale organizations for a decade, and we’ve taken those learnings and best practices and uniquely applied them to your world.

Case in point…Originally, ESI focused on helping hyperscale customers optimize for the workload. These optimization efforts (saving a few dollars or watts on each server) added up quickly for organizations with 100,000 servers.

Then we transitioned to optimizing at the rack. We innovated around shared power, shared cooling, open management, open APIs, etc. After the rack, we optimized the data center. We started aggregating several thousand servers into one shippable solution that provided structure, power distribution and cooling in one pre-manufactured package – the modular data center (MDC). Up until three years ago, the companies using MDCs in the industry were getting 15-20+ racks at a time.

Then the edge changed it all. We started designing micro MDCs. Carriers have thousands of locations that each have a half rack, a full rack or maybe 2-3 racks. Modular data centers were too large – but the basic principles worked well at a smaller scale. We helped carriers, using the same methodology to drive density where it makes sense, to drive the best operational efficiency and time to production — and packaged it into a smaller footprint that makes sense.

But don’t misunderstand – just because we have a standard philosophy around optimizing servers/racks/data center/edge design – doesn’t mean that we think one size fits all. No two locations/no two usage models/no two environments are alike. Dell EMC ESI exists to solve complex problems. We offer tailored solutions for your unique environment. Contact us if you’d like to learn more: ESI@dell.com.

[i]   https://www.wired.com/2016/12/the-end-of-telcos/

About the Author: Jyeh Gan

16 year Dell veteran in global leadership roles ranging from engineering to marketing to strategy. Helped start Data Center Solutions (DCS), led Dell Data Center Infrastructure organization and led teams on assignment in Asia.