The question “what is Linaro?” usually comes shortly after “what company do you work for?” I have a variety of answers honed over the years. I usually ask the Arm question, that is, “Do you know Arm?”. Time was this got blank looks but, lately, most people know about Arm, or at least the Arm processor in their mobile phones. If not, I tell them about Arm’s IP model and how they license technology and how their partners bundle that technology with their own to build devices. Some know, but the majority don’t, that their mobile has a Qualcomm chip in it or that that chip has Arm IP in it, in the form of the CPU architecture, fabric and peripherals. What most do not know is that Arm is everywhere from your car, your washing machine, your camera, your router, your printer and so on. You get the drift. Ubiquitous.
With hardware out of the way, I can talk about software. Again, this is much easier than it used to be. People know about Android so it’s easier to answer and talk about open source and how it works. So far, all I’ve managed to do by this point in the conversation is, if I’m lucky, to describe Arm’s Intellectual Property (IP) model and the rise and rise of open source. I might also mention that the biggest part of my career has been at the mixing point of Arm IP and open source; a combination of strong evolutionary forces that have, literally, revolutionized all our lives. Where would social media be without mobile phones?
What I haven’t managed to do so far is to describe what Linaro is and what it does, but I’m getting there, albeit slowly. The thing is, there’s an awful lot of software that needs to be worked on in order to support hardware in a given industry ecosystem. Millions of lines, I say. Even with fierce competitors (I once likened the Arm partnership to a bunch of pirates), there is a need to collaborate on common software and frameworks. The collaboration is done by a mixture of Linaro engineers (paid for by our membership fees) and engineers from the members. In Linaro’s early days, it was more a question of tidying up the upstream code bases so that the various Arm based SoCs could benefit from features, bug fixes and security updates. This ‘work upstream’ mantra is still true but is now widely accepted in the Arm ecosystem. It’s just how software is developed. Most of Linaro’s efforts these days are in opening up various industry ecosystems to the Arm partnership.
What do I mean by ‘ecosystem’? As an example, Android is an ecosystem. These ecosystems are represented in Linaro by Segment Groups. Linaro currently has four – the Consumer Group (which represents all things Android), the Datacenter & Cloud Group, the Edge & Fog Computing Group and the IoT & Embedded Group. We are in the process of establishing two more, one focused on Artificial Intelligence and the other Automotive. Each Linaro group is supported by members and representatives from that ecosystem. One of the first things the committee establishes is the scope of the group, just what it is that the group is trying to achieve. Each group also decides what software is important to the Arm ecosystem and what software frameworks need to exist in order to properly support the wealth of Arm based hardware. A good example of this is the work that is happening in artificial intelligence as the newly group works to establish frameworks allowing many different types of hardware acceleration from FPGAs to custom Neural Network acceleration hardware.
Today, Linaro is where the Arm partnership collaborates strategically on ecosystems that are important to it, and engineers solutions that benefit Linaro’s membership in particular, and the Arm ecosystem in general. Assuming that my mythical interlocutor is still there, he or she might suggest that Linaro is like the Linux Foundation. My answer is ‘not really’. The Linux Foundation hosts open source projects, allowing companies to help set up traditional open source maintainerships and so on. This is valuable, but Linaro is actually about coordinated, collaborative engineering across a range of, projects. Whilst what we do is good for the software projects that we get involved with, ultimately it also benefits our members as they release products.