On The Future of iOS and Android
We’ve entered the age of iOS and Android penetrating beyond mobile, and there’s a lot happening with respect to how the entire ecosystem is developing. Mobile advancements / investments are now driving the progression of all converged hardware / software. Hardware innovation is also accelerating because of this tailwind. Here are some thoughts on what’s next for the dominant platforms as they stratify in mobile and non-mobile platforms.
- This recent IDC report shows Android pulling away in mobile share worldwide, but in the US Asymco shows it’s a different story, with Android potentially having peaked and in decline. What will happen when the low cost iPhone 5C comes out? It could really shake things up globally.
- There’s more to the platform wars than mobile – Android is starting to take off in non-mobile markets in a massive way – Internet of Things, Television (Chromecast), etc. To date Linux has been the dominant OS but Android is now taking some embedded designs which would have run Linux1. The effective decoupling of Android from carriers for non-mobile markets + the richness of tools and the existing developer ecosystem will likely cement Android as the definitive open source OS of the next decade. This will have pluses for Google but also unintended consequences.
- For non-mobile (plugged in) devices Android + ARM is “good enough” and will win much of the market globally in terms of embedded OS’s (outside of closed / proprietary). Intel will not be able to penetrate unless the device needs a lot of horsepower (e.g. gaming platforms or STB). Why? The entire BOM of Chromecast is $20 and Intel simply can’t price low enough to compete while preserving margin. Meanwhile, Apple will continue to print profits with vertically integrated non-mobile devices (e.g. iTV, iWatch)
- For mobile devices (battery-operated) it’s a totally different story – mobile computing is not about being good enough overall, it’s about being good enough per watt of power consumption. Integrated HW/SW from Apple will likely retain a performance edge—they will be constantly optimizing multi-core chips for performance per watt and tweaking software like they have on iOS72. Who knows if Apple will be 6 months ahead or 18 months ahead at any given time, but it’s a good bet they will be in this range. Low power needs will drive convergence in design.
- Samsung is rumored to be switching to an architectural ARM license because their SoC chips need to compete with Apple’s A6/A7 and Qualcomm Krait (both custom designed cores). Outside of Apple and Samsung (who could be 100% vertically integrated soon), Qualcomm will use custom cores to keep pushing performance per watt up so that it wins the remainder of high end OEM sockets. Mediatek will own the mid and low end. This leaves limited room for Broadcom and nVidia. And none for Intel.
- Fragmentation continues to crop up on Android in weird ways for devs but now is going to the next level—affecting the ecosystem. Tomorrow Apple is no longer just about integrated hardware / software, it’s about an entire integrated ecosystem. The software is built in advance while hardware is designed and spec’ed, developers have your APIs and it all just works together in tandem. To some extent Google can do this with products it controls and distributes like Glass and Chromecast (via OTA updates), but fragmentation will start to hurt Android more and more in smartphone web services which it doesn’t develop.
- New APIs exposed in iOS7 (iBeacon, AirDrop) for local networking will be huge—these are US-centric use cases in the beginning (some Europe and Japan) and will cement Apple at the high end of the market in the US. When iOS7 launches all APIs are backwards compatible 2 years (iPhone 4S and up). This means basically 95% of Apple users will be able to share files with each other (AirDrop) and make payments using Bluetooth / iBeacon, whereas Android will reach round 2 of its fragmentation battle—at the service layer. Only 30% of people upgrade (or are able to upgrade) to the latest Android flavor 1 year later. Devs won’t build networking / payment services in Android apps for BTE and local WiFi sharing because only a minority of Android phones will support for the next 18 months. There are some aftermarket apps in Google Play but compatibility is a nightmare. Again, just to reiterate—yes Android 4.3 adds low power Bluetooth support, but since only about 30% of devices will be running this in a year it effectively slows adoption by 12-24 months behind iOS.
- NFC is dead—that’s not the interesting part though, it’s how Apple was able to replicate NFC functionality with Bluetooth 4.0 and WiFi (they’re also using GPS like Bump did for authentication) and how they standardized all of this into iBeacon in iOS7. While supporting it all backward compatibly to iPhone 4S. A two year old phone upgraded with iOS7 will just work… Bluetooth has arrived – it’s been around forever, but up to now it’s been crappy. Bluetooth LE (also called Bluetooth Smart) changes everything. Connections, pairing, device management etc will finally work 100% of the time, and Bluetooth will be a completely bulletproof, consumer ready, industry leading technology. There will truly be a radio in everything around us and it’s going to enable incredible experiences in mobile. Apple’s iWatch will work so well with your iPhone out of the gate when it’s launched you will be blown away.
- It’s not even funny how bad fragmentation will hurt Android and Google in location based sharing and payments apps, short range sharing, and the type of things developers build on top of iBeacon (e.g. payments). Fragmentation doesn’t matter as much when you are the only one person affected, people deal with it. But when your Android phone won’t communicate with others or at POS terminals (tablets / iPads) it will be tough to rationalize. Bluetooth LE in Android is happening now, but fragmentation is a deal killer for devs, and this ensures that state of the art apps around local discovery / wireless will rarely support Android3. It’s already happening—Tile has raised about $3M from 50K backers and there will be no Android support (these are tagging devices running Bluetooth LE that help you find lost keys etc).
- The next couple years will be one of the more exciting times for local commerce. Sharing and networking services + native bulletproof APIs will finally enable an ecosystem and new use cases and commerce at the point of sale—e.g. Bluetooth powered folios that you pay with using your phone at the table, or loyalty coupons sent to your iPhone (via iBeacon) when you walk into a store. Square pioneered some of these use-cases with its Square Wallet product and now iOS7 standardizes the entire ecosystem. This will really help drive online to offline commerce and attribution tracking. Android will lag here 12-24 months. This means local commerce on Android vs iOS will mirror e-commerce (Android tablet / phone buyers don’t buy as much as iOS). If this plays out at the high end, smartphone sales will remain Apple’s in the US. Worldwide it will be tougher to tell how these use cases develop.
Overall it’s clear there is a stratification happening at the low and high ends of the market, as well as in different geographies. And because embedded platforms are now influenced almost solely by mobile technologies, everything is changing. This is more evidence that mobile platforms will not likely follow the patterns of computing platforms past.
Everything from all of your home wireless routers to the Nest thermostat to the Makerbot 3D printer to the Canary automated home security device run embedded Linux. In the future these designs are up for grabs for Android. ↩
There are a bunch of new power saving features in iOS7 which required converged hardware/software/service layer development – e.g. in multitasking, new background fetch which shuts down apps and only wakes them during an active networking thread. ↩
For some reason fragmentation always elicits the “too many screen sizes” example—but issues such as those can be overcome with responsive design. The real trouble comes in different ways – e.g. a friend I know has been seeing a lifecycle bug crash the keyboards on ALL Samsung phones. It’s not an Android version issue, it crashes the app across revs; it’s something non-standard Samsung is doing to Android itself. These are the types of problems that slow down developers and cause them to reevaluate Android support next time. ↩