Within Geekbench 6, the new A17 Pro chip in the Apple iPhone 15 Pro achieves particularly good results and comes close to the best models from Intel and AMD in terms of performance per core.
Although the iPhone 15 Pro (Max) will only be available from September 22, the first benchmarks are already appearing online. in Jake Punch 6, a CPU-focused synthetic benchmark, the A17 Pro scores 10.33 percent better than the A16 Bionic on a single CPU core. This gain is mainly due to the shift from five to three nanometers.
If we look at the performance gains across all cores, it’s disappointing. Apple makes just three percent profit here.
| to set | single core | Multi-core | |
| Apple A17 Pro | 2P+4E, max. 3.77GHz | 2914 | 7199 |
| Apple A16 Bionic | 2P+4E, max. 3.46GHz | 2641 | 6989 |
| Qualcomm Snapdragon 8 2nd generation | 5P+3E, max. 3.19GHz | 2044 | 5459 |
| Intel Core i9-13900K | 8P+16E/32T, max 5.80GHz | 3212 | 22458 |
| AMD | 16P/32T, max 5.70GHz | 3168 | 22042 |
What’s particularly striking is that with the A17 Pro, Apple is getting very close to the top models from Intel and AMD in terms of performance per core. If we look at the multi-core score, both manufacturers leave Apple far behind. With six cores, two performance cores, and four efficiency cores, you can’t beat chips that have four to eight times as many cores.
Switch to 3 nm
It’s important to stress that we can’t draw conclusions from a single benchmark, but it seems to be true that CPU performance per core has increased relative to the two performance cores of the A17 Pro. The gains Apple is making in switching to 3nm technology look significant. We’ll only know if the switch will also benefit the battery when the device is available.
Read also
The Apple iPhone 15 Pro Max has a faster A17 chip, USB-C, and a 5x zoom lens
Traditionally, moving to a smaller process, 3nm in this case, involves downsizing the architecture without much focus on performance improvements. A significant speed increase is achieved in the second year when the manufacturer has gained an additional year of experience in the 3nm field. We will not know until this time next year how much progress has been made.
