Over the weekend, a Chinese leaker quantified the expected performance of Apple's upcoming flagship chip: 18 percent faster and 30 percent more efficient compared to the A19 Pro. Both figures closely match the specifications published by manufacturer TSMC itself.
The A20 Pro will power the iPhone 18 Pro and iPhone Ultra in September and is Apple's first chip manufactured using TSMC's 2-nanometer process. Concrete performance figures were previously unavailable – a leaked image at the end of June merely hinted at a performance boost. Now, percentage figures are available for the first time. However, none of this has been officially confirmed, and the figures raise a question of their own.
Key Facts at a Glance
- The Weibo account Fixed Focus Digital, citing supplier sources, claims 18 percent more power and 30 percent better efficiency.
- The A20 Pro is Apple's first chip manufactured using TSMC's 2-nanometer process, and it also features a new packaging process called WMCM.
- TSMC itself states that this process will result in either 10 to 15 percent more speed or 25 to 30 percent less fuel consumption – either, not both.
- TSMC specifies exactly 18 percent for the second expansion stage of the process.
- The same source has been wrong several times in the past, most recently regarding the displays of the iPhone 17 and iPhone Air.
What the leak actually says
The information comes from Fixed Focus Digital, a Weibo account that claims to have contacts in the supply chain. Compared to the A19 Pro, the new chip is said to be 18 percent faster and 30 percent more efficient.
A breakdown is missing. Apple itself usually reports the performance gains separately for the main processor, graphics unit, and Neural Engine, and these values are rarely the same. Therefore, a single percentage for the entire chip is a simplification whose point of reference remains unclear.
Why the numbers seem familiar
It's worth comparing this with the manufacturer's specifications. TSMC officially states that its N2 process offers 10 to 15 percent more performance at the same power consumption compared to the previous N3E process – or 25 to 30 percent less power consumption at the same performance. These two figures are alternatives, not a package deal: A chip can translate the available performance into either speed or runtime, but not both completely.
For the second expansion stage N2P, which will go into series production from the second half of 2026, TSMC cites around 18 percent more power and about 36 percent less consumption compared to N3E.
| Info | Performance | Efficiency |
|---|---|---|
| A20 Pro leak | +18 % | +30 % |
| TSMC N2 versus N3E | +10 to 15 % | 25 to 30 % less consumption |
| TSMC N2P versus N3E | approximately +18 % | approximately 36 % less consumption |
The stated 18 percent exactly matches TSMC's figure for the second expansion stage. This allows for two interpretations: Either Apple is already manufacturing the A20 Pro using N2P technology – or the figure does not originate from the supply chain, but from a publicly available datasheet.
Our assessment is therefore cautious: A value that exactly matches a published manufacturer's specification is weaker as evidence than one that deviates from it. Both remain possible.
What else could Apple contribute?
The argument against relying solely on the datasheet is that Apple doesn't only benefit from the manufacturing process. The A19 Pro runs on an improved 3-nanometer process, not the N3E process that TSMC is comparing it to – so the actual difference is smaller than the table suggests.
In addition, there's the new WMCM packaging process, in which multiple components are assembled directly on the wafer. Shorter paths between components improve both speed and power consumption, regardless of the manufacturing process. And the architecture of the processing cores changes with each generation anyway.
Both of these factors together could explain why a chip gains in both dimensions instead of having to choose one. This is not proven.
How reliable the source is
The account has a mixed record. It was correct about the name of the iPhone 16e, but wrong about its sales figures. For the iPhone 17 and iPhone Air, it predicted a fixed 120Hz display without an adaptive refresh rate – the opposite was delivered.
The same source also commented on the launch of the foldable iPhone over the weekend, contradicting a report about a US-only launch. Anyone reading both statements together should evaluate them equally.
What of this matters to you
The more practically relevant metric is efficiency, not speed. An 18 percent increase in processing power is hardly noticeable in everyday use – the limiting factor for current iPhones is rarely the processor. A significantly more energy-efficient processing unit, on the other hand, noticeably extends battery life, especially in conjunction with the speed improvements that Apple has announced for iOS 27.
The situation will become more reliable in September. Apple will present its own figures on stage, but these should be interpreted with caution: depending on the slide, the comparison is sometimes against the direct predecessor, sometimes against a three-year-old model. Independent measurements will only be available with the first test devices.
Is 18 percent more processing power worth upgrading to you, or would longer battery life be the real selling point? Let us know in the comments what you're looking for in your next iPhone.
- iPhone Ultra: Contradictions at market launch, doubts about the name
- iPhone Ultra: Report expects staggered market launch
- iPhone 18: Smaller Dynamic Island already in the base model
