This week’s OpenInfra Live episode features guests discussing the challenges around global connectivity and how open source is aiming to connect billions worldwide.


How different would your daily life be without constant, reliable high speed access to the Internet? This isn’t just a problem for remote, rural areas. It’s a widespread problem affecting over 3 billion people worldwide and 23 million people in the United States alone. 

In today’s episode of OpenInfra Live—the weekly, interactive series hosted by the OpenInfra Foundation—Martin Casado, Bruce Davie, and Amar Padmanabhan joined Jonathan Bryce and Mark Collier to chat about global connectivity and the technology strides that are closing the digital divide.  

How did we get here? 

“The internet came very quickly and it changed everybody’s lives,” said Martin Casado, VC partner at Andreeson Horowitz. “It’s almost a necessity, but because of how rapid the growth was, it has not been homogenous in that growth.” 

Like Casado said, this disparity comes at a time when the people rely on the internet. Amar Padmanabhan, software engineer at Facebook, discussed how the internet has become a crucial tool for improving livelihoods: 68% of people develop new job related skills, 63% improve career prospects, and 61% pursue an education.

If you connect 10 people to the internet, 1 person moves above the poverty line,” he said. “It’s not a luxury—it’s a necessity.” 

Network evolutions have always been in service of new computing platforms. When asked what is driving the network changes, Bruce Davie talked about the evolution of compute architectures and how that’s impacted networking requirements. 

“One of the biggest shifts in networking was the introduction of smartphones – now, there is a computer in everyone’s pocket,” Davie said. “We’ve gone from an architecture that’s connected a dozen computers to a billion computers; it’s quite the shift.”  

Davie further explained how decentralizing the compute geographically was something that networks were still evolving to meet. “Where you connect to the network could be anywhere on the planet instead of in a centralized lab,” he elaborated. “The architecture for connecting billions of mobile devices is still an evolution.”  

As networks have evolved, there are two parallel architectures and they are starting to converge—one for the internet and one for mobile devices. 

Enter Magma.

Magma is an open source packet core that’s taking an internet-centric approach and applying that at the edge, says Davie. He says this is what it will take to connect devices around the world. 

After all, 50 billion devices will be connected to the network by 2030. Casado predicts that this scale is already a reality now. 

As more and more devices are added at the edge in a cost-effective, low latency way, Davie and Collier talk about how compute requirements increase. Take autonomous vehicles for example. This is an emerging use case where it’s obvious how quickly you need to get information to the devices in the right time frame. 

“Network speed increases driving the need for more compute and vice versa,” said Collier. “Intelligence and the AI/ML use case changes everything when you have super low latency more so than the raw speeds, but that’s not what’s disrupting the architecture.” 

This is why you see a cloud computing system at the edge when you look at the architecture for Magma. 

“When you look at Magma, it terminates the wireless protocols at the edge so you can build a core that is independent of what’s going on,” said Davie. “It means you can create a more scalable, reliable core and if I can get access to any radio, I can stick it on the Magma core so I can have a common edge system that terminates the radios at the edge and provides common core services to the radio devices. 

Casado jumped in to encourage the group to look at the bigger picture, that the need for Magma highlights the global need for connectivity. Bryce connects this to the commercial forces that are driving this network evolution. 

Bryce referenced a recent Light Reading article that said the number of customers for wireless internet service providers (WISPs) is doubling every five years, growing at an annual rate of 73%.

“Last year, it was already a $4.4 billion market in the US for revenue in this category which has become a use case for Magma,” he said. “Part of the reason that Magma as an open source project can exist is because there is now a commercial opportunity and a need and these WISPs are filling some of those spaces in the US that have not had great coverage and going after markets that are not appealing to traditional telco operators.” 

Padmanabhan said the community broadband radio spectrum (CBRS) unlocked the spectrum barrier, opening access for innovation. Even with previous networking use cases like SDN, it was never optimized for a particular application. As new use cases emerge—like wearable technology—you need to figure out how to create a consistent runtime between the edge and the network. 

“The open platform that is best suited for heterogeneity is going to be in the best position to unlock the next billion dollar use case,” said Padmanabhan. 

This opportunity is one that OpenStack took over the past decade with telco adoption for software defined networking capabilities. 

“Over 4 billion people are connected to networks powered by OpenStack,” said Collier. “Now, nine out of 10 of the top providers in the world are powered by OpenStack. When you design a technology or enter a market with different tools that enable people to do things, they will do things you would never expect.” 

Hear more about the open source opportunities to advance global connectivity by listening to the rest of the discussion. 

If you want to help us build the network to connect the next billion people, join the OpenInfra Foundation and get involved in the OpenStack and Magma communities.

Allison Price