All the ways iOS 13 is going to be faster than iOS 12

Apple announced a slew of new features coming to iOS 13 at WWDC. But you probably shouldn’t worry about all those fancy new capabilities slowing your iPhone to a crawl.

In fact, Apple Senior VP Craig Federighi opened the iOS 13 section of the WWDC keynote by saying “nothing is more important to our iPhone users than performance. So this year, we worked top-to-bottom making everything faster that you do the most.”

Here are the ways you can expect your iPhone to get even faster when iOS 13 is released this fall.

Faster Face ID

Apple

Face ID unlock is already fast, but Apple is making it up to 30 percent faster. That slight pause as you wait for your phone to unlock could be almost imperceptible. There’s not much more to say about this one–Face ID is still what it is, and iOS 13 doesn’t make major changes to how it’s used, its accuracy, or its range or field of view. Those sorts of changes will probably require a second-generation TrueDepth sensor (which may come in a future iPhone).

Smaller app downloads, faster app launching

Apple

With the release of iOS 13, Apple is changing the way apps in the App Store are packaged together. The result, it says, is app downloads that are up to 50 percent smaller and app updates up to 60 percent smaller.

That means faster downloads, less strain on your mobile data plan, and less storage space taken up by apps. (Frankly, though, app data like photos and music are still going to eat up most of your iPhone storage space.)

But it has a secondary effect of making apps launch up to twice as fast. Launching an app means reading it from storage, decompressing the compressed bits, and storing the active parts in RAM. When the app is half the size, all of that can happen up to twice as fast.

We should note that you’re unlikely to see this performance boost during the iOS 13 beta period. The new app packaging method won’t reach users until Apple starts accepting App Store submissions that target iOS 13, usually right around the time iOS 13 launches. So even if you have the iOS 13 beta, don’t expect smaller apps and faster launching until the iOS 13 release this fall.

And lots of under-the-hood stuff

Smaller, faster-launching apps and faster Face ID are concrete things that any user can wrap their head around. But iOS 13 is getting faster in a whole bunch of other esoteric, under-the-hood ways that will work in tandem to make a lot of apps and operations faster, smoother, or more responsive. Some may improve battery life, too.

Apple

At WWDC, Apple showed a slide listing a lot of these under-the-hood changes. It’s a big list, but probably not comprehensive:

• Improved kernel thread scheduler
• DLYD3 closures
• System frameworks
• More efficient shared cache
• Improved background process efficiency
• Compiler improvements
• Cached system state
• iCloud optimizations
• Asynchronous Metal shaders
• Swift code optimization
• Dynamic network switching
• Reduced memory usage on Super Retina displays
• Auto layout
• High-efficiency memory allocation
• APFS cache delete
• Optimized FairPlay decryption
• Launchtime daemon throttling
• Prioritized pageouts

Keep your expectations in check

It bears repeating that you can’t expect iOS 13 beta releases to feature all the same performance characteristics of the final release. Beta releases are not fully optimized and often have bug-tracking and telemetry code and the like, which affects performance.

We should also mention that you shouldn’t necessarily expect iOS 13 to deliver meaningfully better scores in benchmark tests like Geekbench. Once an application like that is loaded, running through its own algorithms to test processor performance, it’s beyond the reach of most of these new iOS optimizations. These iOS optimizations impact performance characteristics that we can feel when load apps, play media, and scroll through social media feeds, but often have little to do with running highly optimized CPU-bashing tasks that run in the background without human interaction. Geekbench and its ilk are valuable benchmarks, but they stress theoretical processor performance more than operating system limitations.