SAN FRANCISCO -- Docker has inked a partnership with ARM to better enable developers to build applications for cloud, IoT and edge computing systems on the RISC architecture.
ARM's dominance in edge computing, mobile and IoT make it an attractive collaborator for Docker, as the container company looks to expand its reach. Docker on ARM is particularly important as automated software updates on edge devices becomes a bigger trend, said James Governor, co-founder and analyst at RedMonk, a developer-focused analyst firm in Portland, Maine.
"Docker needs to be everywhere, and ARM keeps threatening to make a dent on the server market," Governor said. This integration agreement means millions of ARM developers can now build Docker-based apps for their workloads.
Docker on ARM starts with desktops
The partnership kicks off with integration of ARM capabilities into Docker Desktop Community and Docker Desktop Enterprise AWS EC2 A1 instances, said David Messina, executive vice president of strategic alliances at Docker, ahead of DockerCon 2019 here this week. Docker Desktop, available for Windows and MacOS systems, enables users to build and test containerized applications with their chosen framework and development language.
The companies are making Docker-based tools available to ARM developers as an extension of the ARM Neoverse CPU, optimized for cloud-native server workloads.
Meanwhile, a potential Kubernetes element of the partnership would further enable ARM adoption in the cloud-native world by enabling cloud developers to save costs without sacrificing performance, said Matt Kimball, an analyst at Moor Insights & Strategy in Austin, Texas.
"This can't be overstated; it gives developers a choice -- a choice of architectures; a choice of vendors," he said.
Moreover, it enables the proliferation of ARM across the device-to-data center continuum that spans from the sensor on a factory floor to the edge to the data center, he said.
Support for Docker on ARM is important for not only cloud-native companies that want to run workloads on the fastest and most efficient architecture; the same is true as you move toward edge devices and the end of the network.
You want to be able to run your workloads on the architecture that's going to be a better fit.
Matt KimballAnalyst, Moor Insights & Strategy
"You want to be able to run your workloads on the architecture that's going to be a better fit," Kimball said. "Traditionally, in a lot of those apps, it's been RISC architectures, whether MIPS, PowerPC or in this case, ARM."
The thrust of this Docker-ARM partnership is good in that it indicates a shift toward more virtualization and more containers.
"Let's face it, the x86 [architecture] dominates servers in terms of CPUs, but 100 times more ARM chips ship every year because they're used throughout a bunch of other applications throughout the network, throughout edge devices," said Jim McGregor, founder and analyst at Tirias Research in Mesa, Ariz.
The industry is undergoing a shift toward accelerated computing or heterogeneous support where applications are able to run on not just CPUs, but a GPU, neural processor or some type of other dedicated accelerator.
"Containers allow you to break up those applications and run either a part of the app or the entire app on the resource that's most effective," McGregor said.
Docker and ARM will work together to provide management of the application lifecycle from development to secure device management, unified development environments for heterogeneous computing, and ways to scale cloud workloads and consolidate edge workloads, the companies said.
