MobiledgeX is an edge computing company founded about two years ago by Deutsche Telekom and headquartered in San Francisco. But, while MobiledgeX is owned by DT, it works with mobile operators worldwide, and it considers them to be its business partners.
MobiledgeX said in September that it now counts 13 mobile operators as part of its global edge network initiative. The company first announced its initial deployment earlier this year.
It aims to help operators combine edge resources with their wireless infrastructure, including their 5G rollouts. It’s also building a marketplace of edge services.
At the Open Networking Foundation (ONF) summit in Santa Clara, California, last month MobiledgeX CEO Jason Hoffman said, “No matter who you ask ‘what is edge?’ they always have an answer that is very specific to themselves. But, it involves everybody.”
Those players include mobile operators, tower companies, cable providers, infrastructure vendors, governments, site owners, and enterprises.
Hoffman talked about some of the complexities of building edge infrastructure for multiple operators. He said, “You do start hitting a boundary of a regulatory regime that is not present in large public clouds, and he added that “depending on different countries and how 5G is going to happen, will determine how edge happens in that country as well.”
In addition to connecting mobile operators, MobiledgeX also wants to connect all the major public cloud providers. Hoffman said there are “little differences” in the various cloud architectures that will feed into the edge architecture, and those must be accounted for, as well.
Defining the edge
Analyst Chetan Sharma recently penned a report “Edge Computing Framework: Understanding the Opportunity Roadmap” along with support from MobiledgeX. The report notes that there is no singular definition of edge that has gained consensus. “According to ETSI, ‘Mobile edge computing provides an IT service environment and cloud computing capabilities at the edge of the mobile network, within the radio access network (RAN) and in close proximity to mobile subscribers,’” states the report.
But, it adds that “IEEE defines edge computing as something that places applications, data, and processing at the logical extremes of a network rather than centralizing them. Linux Foundation says something similar – ‘The delivery of computing capabilities to the logical extremes of a network in order to improve the performance, operating cost, and reliability of applications and services.’”
Despite disagreement among various edge players, Sharma’s report lays out six basic edge capabilities:
- Compute – Providing a compute platform for data processing, content processing, and image processing.
- Access to Network Intelligence – Access to immutable network data for key variables such as location, identity, traffic, geospatial, and security.
- Latency – Dropping latency to a few milliseconds.
- AI/Analytics – Availability of AI algorithms that can work independently or in concert with a distributed edge or distributed cloud architecture.
- Storage – Storage location for data and content.
- Data Residency - There are multiple scenarios whether driven by regulations (e.g. GDPR) or security (IoT Firewalls) or functionality (e.g. in agriculture) where the data can’t travel too far from the source that produces it.
Edge and the future
As far as the edge infrastructure, Hoffman said, “Clouds exist in the hundreds, but edge needs to exist in the hundreds of thousands of base station deployments.” And he added, “Clouds give the impression of infinite resources and scale; while edges are finite in size, and there’s always a scarcity model.”
Hoffman is a bit of a futurist. At the ONF event he talked about the changing way consumers interact with the internet. In the past, it’s been a “heads-down” experience where people looked at their computer monitors or their phones. But in the future a much larger percentage of traffic will be more of a “heads-up” interaction where consumers will talk to their devices, while they’re on the move.
And this heads-up model is much more complex than the old paradigm. Hoffman said in the beginning of the internet, a static client talked to a static server. Then smartphones came along, and they’re mobile, but the backends could still be static. Now, the model has shifted where even the backend can move around as consumers travel from one country to another.
“Devices and their apps will trigger the consumption of edge,” said Hoffman. “Edge workloads have a highly precise and accurate GPS location. Edge workloads have to have an understanding of economics and sovereignty.”
In addition, Hoffman predicts that people will demand more personalized behavior from their mobile devices. He made it sound as if we’ll all wear our personal mobile devices on our faces, perhaps via some kind of glasses, and our devices will use AI to provide information and communication, almost like a second brain.
This will require a lot of edge computing in the future. Hoffman said the edge will be “basically a trading engine,” and it will “need to be more privacy secure than the overall internet has been.”