Qualcomm's Take on Open Mobile Platforms

This week at LinuxCon, the Qualcomm Innovation Center (QuIC) joined The Linux Foundation at the platinum level. This represents Qualcomm’s strong commitment to Linux and to collaborative development. QuIC President Rob Chandhok, who keynoted at LinuxCon, also shared with us his ideas around mobile platforms and openness.

Can you tell us about Qualcomm’s take on the Mobile Linux marketplace?‚Ä®

Chandhok: Given the fragmented nature of the mobile OS world, Qualcomm believes the greatest critical mass and potential for significant developer leverage is mobile Linux. Unfortunately, mobile Linux mirrors the larger fragmentation we see in mobile operating systems. That fragmentation is largely driven by individual companies seeking competitive advantage via differentiation. These are legitimate business goals. However, we believe that these goals can be met while also establishing a common mobile Linux upstream baseline upon which differentiating capacities are built by individual players. In short, it is contrary to the interests of mobile developers for fragmentation to happen too low in the software stack. To the degree that this fragmentation is harming the business interests of mobile developers, it also is harming the interests of the rest of the players — OS distributions, technology providers, device makers, wireless operators and, ultimately, users.  Qualcomm is seeking to work with all the critical industry players to move towards the definition of a common baseline package.‚Ä®‚Ä®

The Qualcomm Innovation Center appears to be a good example of how companies can proactively ensure their products integrate with mobile open source projects. Can you tell us more about the Center and the importance of supporting upstream projects?

Chandhok: The Qualcomm Innovation Center, Inc. (QuIC) exists as an independent Qualcomm subsidiary to enable us to make upstream contributions to mobile open source efforts (not just those based on Linux, but also in the area of web technologies) such that open source High Level Operating Systems (HLOSes) and web technologies incorporate features that can enable the maximum optimal leverage of the advanced features and capabilities of Qualcomm chipsets. As such, it serves Qualcomm’s business strategies, obviously. But it also directly serves the interests of mobile developers, device makers, wireless operators and, it is important to always remember, end users. This is because a significant number of smartphone models from device makers around the world are powered by Qualcomm’s Snapdragon™  family of chipsets. Since a very large fraction of these models are running an open source operating systems (OS), working to make sure those OSes have the hooks necessary to optimally leverage our chipsets helps developers (addressable market for premium apps), device makers (get the most out of our chipsets), operators (offering advanced devices and services) and end users (coolest and most useful devices and services). Let me add that many of the contributions we are making upstream are also useful additions to the overall code base generally — even if the actual commercial device is not based on a Qualcomm chipset.‚Ä®‚Ä®

You recently posted a job opening for a Linux kernel software engineer. How will this position help to advance collaboration between Qualcomm and the Linux development community?‚Ä®

Chandhok: The Linux kernel is the central key element of the common baseline package we would like to see emerge in mobile Linux. This engineer will help QuIC make important upstream contributions to the Linux kernel that, ideally, would be part of this common package. At a minimum, we expect this individual to help make valuable upstream kernel contributions that make it into the main distributions on offer today, or coming in subsequent years.



How important is a common set of tools for delivering consistent support for Linux-based products to your customers? What are you doing today to ensure you can achieve that?‚Ä®

Chandhok: Well, every developer would love to have only one set of really excellent tools, right? In that utopian construct, the simple answer to your questions is: Very Important. But a single common set of tools is not going to happen. We are happy to help tools developers improve their tools but, at the end of the day, the best place we can have an impact is by providing a common set of excellent tools specific to the particular features and capabilities of our chipset system software, tools which integrate well with the larger set of tools being provided to the developer and manufacturer communities. We have a long ways to go on our tools development efforts. I wish we were farther along. But we are getting there. I would point to the Adreno Profiler, which helps developers write for our GPU; the Snapdragon Mobile Developer Platform (MDP), which we announced at Uplinq 2010 at the end of June, and the Snapdragon MDP’s Trepn™ power profiling tool that helps developers optimize their app’s power-usage. And more tools are in the works.  ‚Ä®‚Ä®

Having been founded in 1985, Qualcomm has been around the block. What knowledge and experiences are you drawing on to advance the company in an environment now dominated by open platforms?‚Ä®

Chandhok: Let me start by saying quite simply: We really like the move to open platforms. The emergence of open mobile platforms is essentially an organic evolution towards standardization.  Standards, as nettlesome as defining them can be, are good things. Our primary differentiation is in our advanced chipset+system software designs. That is where we compete. Open platforms pose no issues for our business model. In fact, they are helping us by creating a larger set of cooperative relationships for us. Qualcomm working as a good partner with many companies across the mobile space, including those we also compete with, is a good thing — for us and for the industry. The direct answer to your question can be found in my previous answers, but let me sum it up:  We are drawing on our deep knowledge of all things mobile (a knowledge and expertise which I would argue is second to none) to bring optimal uniquely mobile capabilities to mobile open source. And, for mobile developers and our manufacturer customers, we tightly integrate and optimize open source OSes (all the leading OSes our customers are implementing, actually) with our chipsets, which themselves feature software that functions as a highly optimized system. The intent quite plainly is so HLOSes–and the apps that developers write to them–run best on Qualcomm chipset-based mobile devices.