Altran and CableLabs have teamed up on "Project Adrenaline," an open source initiative that aims to help the cable industry build and manage edge networks and smooth the path for apps that can run on them.

And while Adrenaline is initially focused on cable, the broader aim is to apply the resulting open source platform to multiple industries while still staying aligned with Kubernetes.

"It's a cable-first initiative, but this is also a general purpose platform," said Shamik Mishra, VP of research and innovation at Altran, a company that's primarily focused on engineering and R&D services for multiple industries, including the telecom sector.

"We're not trying to constrain what use cases there are [for Adrenaline]," added Randy Levensalor, principal architect with CableLabs's future infrastructure group in the office of the CTO.

The initial use cases being explored already span beyond just cable's core hybrid fiber/coax (HFC) network. In addition to helping to accelerate work in areas such as a virtualized converged cable access platform (vCCAP) core, Adrenaline is focused on virtualizating mobile cores as mobile and wireless services become increasingly important to cable operator strategies, as well as security and online gaming.

"We want to make it easy to manage those platforms, so you can do any of these use cases," Levensalor said, noting that about half of CableLabs's members are also mobile operators.

Altran's Mishra believes Adrenaline will help to build out an edge computing ecosystem that covers not just the placement of the compute resources closer to end users, but also how new applications are developed to run on these new low-latency networks. Most developer ecosystems, he adds, don't have the time and money to go through several months of onboarding procedures for a simple application that, for example, could be focused on gaming. And while applications can run very fast on powerful cloud servers, edge networks aren't necessarily equipped or designed to run low-latency gaming, AI or virtual reality applications.

"There's definitely a need in the market to simplify the deployment aspects of developer apparitions, which, I think, has not been a focus too much in the edge computing world," Mishra said. "The latency benefit through network placement actually gets lost because the compute itself takes a lot of time."

Pumping Adrenaline in 2020
Mishra said the near-term plan is to open source Adrenaline in the first half of 2020 and to explore the first use cases that can be placed on top of the platform, with virtual CCAP or a virtualized mobile network among the possible first ones that will be pursued.

He also expects the project to produce some reference architectures for both access and core networks, followed by some early trials and pilots, later this year.

Altran and CableLabs are running point on the project, but there's work underway to extend that outreach and collaboration to additional vendors, Levensalor said.

And they're making a point to work with other open source projects that fall within the scope of edge computing and cloud native designs. "We'll work this year to make sure everything we release through Adrenaline is aligning with the other work going on in the broader Kubernetes community," Mark Bridges, VP of CableLabs's future infrastructure group, explained.

Project Adrenaline evolves out of CableLabs's recent work with field-programmable gate arrays (FPGAs) and graphics processing units (GPUs) as hardware accelerators.

As Levensalor explains in this blog post about the project, frameworks such as OpenCL attempt to "bridge the gap" so apps can be written across CPUs, GPUs and FPGAs. "Unfortunately, this interoperability comes with a performance cost that makes these frameworks a poor choice for certain edge workloads," he added.

Related posts:

— Jeff Baumgartner, Senior Editor, Light Reading