Does a new report say how powerful or power-efficient Apple’s rumored “M2” chip will be? The generational transition of Apple Silicon chips in the past can provide hints.
Apple is basically expected to debut the “M2” Mac chip later in 2022. While there are some rumors about the upgrades it will bring to the Mac, MacWorld’s Cross believes that by studying Apple’s iPhone’s A-series chips, it’s possible to get a clearer picture of just how far the “M2” will be.
The simple meaning of the report is that the M2 chip versus the M1 will be similar to the A15 Bionic versus the A14 Bionic. From there, he deduced what “M2” might look like on a similar jump basis.
“The M1 builds on the basic architecture of the A14 and expands the number of high-performance CPU cores (four instead of two) and the number of GPU cores (eight instead of four). For lack of a better explanation, it may be quiet For the ‘A14X’, it will also appear in the iPad Pro .”
Basically, the argument is that next-generation chips will have double the number of high-performance cores and double the number of GPU cores.
“When trying to predict the capabilities and performance of the M2, we’re going to base ourselves on the assumption that it will follow a similar pattern, only this time it’s the A15: the same architecture, with twice as many high-performance cores and GPU cores,” Cross continued.
Based on the A14 and A15, MacWorld then deduced just how much of a performance boost the new “M2” could bring to the Mac. Based on benchmark estimates, that could be a significant jump.
As this article points out, the “M2 Max” scores over 14,000 points, putting it well ahead of Intel’s latest and greatest Alder Lake Core i9. Of course, the “M2 Max” may also be more power-efficient.
In addition, the extrapolated data also shows a significant improvement in graphics performance, including a 3DMark score of 10,409, which is higher than the M1’s score of 7,781.
MacWorld noted that it didn’t account for any performance gains from the improved manufacturing process, as there was no way to extrapolate or quantify what that might look like.
Of course, these estimates are only rough projections. Because of this, there should be a lot of confidence in the estimated scores and gains.
It remains to be seen whether “M2” will perform as well as the inferred data, as the architectures are not the same and the expected use cases are not the same.