Apple M1 Max is the most powerful of the new processors presented by the company during yesterday’s event, from which the new Macs and further hardware innovations from the Cupertino house emerged, and to get an idea of the capabilities offered by this new chip it is enough think that how graphics power may be able to reach the levels of PS5 o Nvidia RTX 2080.
To confirm this statement, mathematical calculations and the first unofficial benchmarks arrived on the Notebookcheck website, which carried out some slightly spannometric calculations pending more in-depth analyzes on the issue, but the interesting data on computational power was still released by Apple . To tell the truth, it is not easy to make such a comparison between SoCs with even profoundly different architectures, but i numbers they are indeed in line with the claim.
Apple has reported that the Apple M1 Max, the SoC with a 32-core GPU inside, can achieve 10,4 teraflops of computational power, with 4096 execution units, 327 gigatexels per second and 164 gigapixels per second as regards the graphics component of the chip.
All this, however, consuming much less power in terms of energy absorption, reaching a very advantageous power / performance ratio (about 100W less than a laptop with a discrete GPU).
To be sure, there are no direct comparisons made through console and GPU benchmarks, but based on the numbers it is possible to make the bold comparison.
Based on these, 10.4 teraflops puts the M1 Max’s GPU in line with, if not slightly above, the results that can be achieved by an Nvidia RTX 2080 GPU (10.07 Teraflops) or PS5 (10.28 Teraflops), which can give an approximate idea of the power that can be reached by the new chip, while maintaining a very low consumption, up to 40% less than a high-end laptop with a discrete GPU.
Just from yesterday’s presentation, the first official information on Apple M1 Pro and M1 Max arrived, which are at the base of the 14 and 16-inch MacBook Pro 2021. The M1 Max chip, in particular, is made up of 57 billion transistors, practically doubling those present in an AMD Navi 21 GPU and all this is also inserted inside a chip from 5nm TSMC, surpassing the amount of transistors present even on the most “dense” chips seen so far at 7nm, in a smaller space.