Predictions of processing and graphics performance for Apple Max M2 and M2 chips show that Apple’s future silicon can provide good results compared to the current Max M1 and M1 Max. The multi-core performance of the M2 Max can surpass that of the Alder Lake Core i9-12900HK. However, its 40-core GPU could theoretically be close to the graphics performance of a RTX 3070 laptop.
Recently, we have heard a lot of news that indicate that Apple intends to introduce devices equipped with M2 chips in March (March 1400 or April 1401). We have not yet seen the performance of the Apple M2 in the real world; But with the recent scale of Apple chips compared to their predecessors, one can get some fair ideas.
Related articles:
Macworld has now come up with numbers that can give us an approximate estimate of the processing and graphics performance of Apple’s M2 and M2 chips. To estimate these numbers, a similar scale was used between Apple’s A14 and A15 bionic chips for the M1 and M2. Accordingly, the number of powerful M2 cores in the multi-cluster structure of the processor cores could be twice as many as in the A15 bionic.
The A15 Bionic offers two powerful cores and four cores for more efficient use and five graphics cores in the iPhone 13 Pro models. Therefore, according to the previous scale, we can consider four powerful cores and four low-consumption cores for Apple M2. Also, considering the M1 scale compared to the M1 Max, it can be assumed that the M2 Max has 6 powerful cores and two low-power cores, and also uses a maximum of 40 graphics cores.
M2 Max performance 7% faster than Core i9-12900HK
The A15 bionic chip is about 7 and 20 percent faster than the A14 bionic in single-core and multi-core geek benchmark tests, respectively. Given the current M1 and M1 Max single-score scores, Macworld estimates that the M2 and M2 Max scores in this test will exceed 1,800.
Because the current M1 Max has more powerful cores than the M2, it is estimated that the M2 chip will score less than 9,000 on the multicore test, or 27% less than the M1 Max multicore scores. However, the M2 Max is expected to be at least 19.5 percent faster than the M1 Max with a score of 14.632. This means that the expected M2 Max could be 2.7% faster than the Core i9-12900HK Alder Lake chip in the multi-core test; While with much less power consumption, it performs the same in single-core testing.
GPU performance of M2 Max chip equal to RTX 3070
In terms of graphics performance, one can expect the M2 to be equipped with 10 graphics cores; But the M2 Max may have up to 40 cores. In simple graphical benchmarks such as 3DMark Sling Shot Extreme Unlimited, the performance of the A15 bionic chip is about 24% better than that of the A14 bionic. Considering the iPad Pro scores in this benchmark, since it does not have a desktop version, it can be predicted that Apple’s M2 chip may gain more than 10,000 points with a 33% increase.
In another benchmark, 3DMark Wildlife Unlimited, which is suitable for cross-platform testing, the A15 performs nearly 30 percent better than the A14. M2 is predicted to be similarly about 30% faster than M1; But the M2 Max is expected to be up to 23% faster than the M1 Max.
According to Macworld, the predicted score for the M2 Max in 3DMark Wildlife Unlimited is very close to the RTX 3070 mobile score; Of course, this does not appear to be the case in the 3DMark database. However, it is not difficult to guess that the M2 Max GPU will work shoulder to shoulder with the advanced desktop graphics made by Nvidia and Imedi and Intel.
However, Apple is still not expected to offer support for any porter interception with the expected chip graphics. In general, the shift from M1 to M2 could be similar to what happened between Apple’s A14 and A15 bionic; So you can expect 20% CPU performance and 30% GPU performance improvement.
Needless to say, predictions cannot be accurate; This is because Apple is expected to build the M2 chip with TSMC’s N4 architecture, which is more energy efficient and can be clocked at a higher operating frequency. According to the TSMC, the four-nanometer manufacturing process should have 11 percent better performance at the same power consumption or 22 percent less power to deliver the same performance as the five-nanometer process.
There are other variables involved in the performance of the chips, such as Apple Neural Engine, video encryption and decoding engines, memory on the cross-section of the die, and the performance of the memory storage system; With all of this in mind, Apple is expected to once again launch high-powered chips with controlled power consumption compared to current flagship laptop processors.
Source link
