iPhone XS Benchmarks
Geekbench 4 is a benchmark that measures overall performance, and no other phone comes close to Apple’s new handsets on this test. The iPhone Xs notched 11,420, and the iPhone Xs Max hit 11,515. The older iPhone X scored 10,357, so that’s about an 11 percent improvement.
The iPhone XS is faster than an iMac Pro on the Speedometer 2.0 JavaScript benchmark. It’s the fastest device I’ve ever tested.
This result doesn’t make much sense to me. The iMac Pro has a higher clock rate and more cores. And it’s inconsistent with the Geekbench Mac and iOS benchmarks.
Update (2018-09-26): Jason Cross:
The iPhone XS and XS Max offer essentially the same performance. Single-threaded CPU performance is about 13 percent faster, very close to Apple’s claimed 15 percent speedup. That helps contribute to a very small improvement in multi-core performance, but since the four energy-efficient cores aren’t really any faster, the difference is minimal.
Geekbench’s GPU test uses Metal to perform computational tasks, so it’s a pretty good indicator of the graphics processor’s ability to do math without actually rendering 3D graphics on your screen. It’s almost 40 percent faster, which is impressive, though not quite the “up to 50 percent” that Apple claims.
Update (2018-10-03): Greg Parker:
ARMv8.3 adds a new float-to-int instruction with errors and out-of-range values handled the way that JavaScript wants. The previous insns to get JavaScript’s semantics were much slower. JavaScript’s numbers are double by default so it needs this conversion a lot.
This could help explain what DHH saw.
Update (2018-10-05): Andrei Frumusanu:
Overall the new A12 Vortex cores and the architectural improvements on the SoC’s memory subsystem give Apple’s new piece of silicon a much higher performance advantage than Apple’s marketing materials promote. The contrast to the best Android SoCs have to offer is extremely stark – both in terms of performance as well as in power efficiency. Apple’s SoCs have better energy efficiency than all recent Android SoCs while having a nearly 2x performance advantage. I wouldn’t be surprised that if we were to normalise for energy used, Apple would have a 3x performance efficiency lead.
Update (2018-10-09): See also: Hacker News:
iPhone XS Max vs Xiaomi Pocophone F1 speed test comparison! What’s gonna happen?!
Spoiler: The “cheap” phone wins.
Apple can have the “3x faster” SOCs but if they don’t translate to real world gains in important tasks like launching apps, what’s the point?
iOS 12 was just a first step. Still needs more performance improvements (app launch & multitasking).
Update (2018-10-12): John Gruber:
Turns out JavaScriptCore (Safari’s JavaScript engine) doesn’t use this new instruction yet — it should make things even faster once it does but the A12 chip is getting these benchmark scores without this new instruction’s help.
Update (2018-11-13): Marco Arment
All of this power in the A12, and I need to throttle Overcast’s Watch-transcoding engine, even when connected to power, because iOS kills any app that uses more than 80% of the CPU over 60 seconds.
As someone who works on machine vision on iOS: Yes! We’ve observed it on all iPhone models since the 6S, a few earlier ones too. iPhone benchmarks are kind of misleading because you can’t use that power for more than a minute or two. iPads seem better, but we work with them less.
If you look at some of Apple’s samples they actually subscribe to thermal notifications and warn you when the phone is getting too hot and will start downclocking.
