At the One More Thing event taking place early November 11, technology giant Apple has officially launched three new Mac products including MacBook Air, MacBook Pro and Mac mini. The most notable point of this new Mac line comes from the fact that all 3 devices use a brand new Apple M1 chip that Apple researches and develops on ARM architecture, instead of Intel chips like the old Mac generation.
Also at the launch event, Apple confidently affirmed that the M1 is the fastest mobile chip in the world today. This statement of Apple later caused a lot of controversy, especially when the company did not publish any specific benchmark results of the Apple M1 chip.
Recently, however, a series of benchmark scores of this chip have been revealed. And the results seem to contradict Apple’s claims.
Is the Apple M1 really more powerful than the mobile chips from AMD and Intel?
Apple’s M1 chip is developed on today’s most advanced 5nm process, which is provided by semiconductor manufacturing company TSMC. For comparison, Intel’s x86 architecture mobile CPU models are built on a 10nm process, while AMD uses a 7nm process.
M1 equipped with 8 cores (including 4 high-performance cores and 4 energy-saving cores), integrated up to 16 billion transistors,
In essence, shrinking the process will give chips the advantage of reducing power consumption, lower temperatures, while also integrating higher transistor densities. However, in the case of the Apple M1 alone, the advantage is almost absent.
According to information posted by the Wccftech site, the Apple M1 chip has been beaten in performance by a series of mobile CPU models from AMD and Intel, based on benchmark results using Cinebench R23 – software commonly used by Use technology expert to evaluate CPU performance. Thanks to the ability to squeeze 100% out of CPU processing and calculation performance, the Cinebench R23 benchmark score is more reliable than Geekbench. The benchmark results themselves with Geekbench (which run many algorithms and take the results averaging) are also prone to deviations if the algorithms are optimized specifically for a certain architecture.
Table comparing multi-core performance by Cinebench R23 software
Specifically, in the Cinebench R33 multi-core performance test, all Ryzen 4000 CPU models on AMD’s 7nm process showed absolute dominance in terms of performance. Apple’s M1 chip is only about 7508 points multi-core, far behind AMD’s most advanced mobile CPU model, the Ryzen 9 4900h (reaching 11061 points – 38.2% difference in multi-core performance).
Notably, a chip built on Intel’s 14nm process the i7-10850H even has multi-core performance almost on par with the M1. Compared with the i7 1185G7 – a 4-core CPU model of the Tiger Lake line developed on Intel’s 10nm process, the M1’s performance is only 1000 points higher, despite Apple’s chip having more than double the multiplier.
Coming to the Cinebench R33 single-core performance test, the M1’s performance was significantly better, surpassing AMD’s Ryzen 4000 series and only losing to Intel’s 2 representatives. This is also quite understandable, as the Ryzen 4000 mobile CPU models are not really powerful in terms of single-core processing, since the IPC of the Zen 2 architecture is still significantly inferior to Intel (this problem has been fixed with Zen 3 architecture & Ryzen 5000 series on PC).
But it must be added that the single-core performance score doesn’t really make much sense, as most current applications take advantage of the CPU’s multi-threaded power.
Table comparing single-core performance by Cinebench R23 software
Based on the aforementioned benchmark results, technology news site Wccftech has clearly confirmed that the Apple M1 is completely not the fastest mobile chip available today. Despite its impressive processing performance, Apple has clearly been overstating that the chip is far ahead of the mobile chips from AMD and Intel.
However, the switch to M1 chips (instead of CPUs from AMD and Intel) is still a reasonable move for Apple at the moment, according to Wccftech site. Thanks to an exclusive contract with TSMC, Apple has had the opportunity to access the latest CPI development today, earlier than its competitors.
In addition, Apple can also take advantage of the 5nm process to combine with its ARM architecture. As a result, Apple’s home-grown ‘home’ chips will be significantly cheaper – helping to increase profit margins, while performance isn’t too inferior to Intel and AMD. In fact, thanks to Apple’s extremely well-optimized hardware and software capabilities, it’s hard to tell the difference in performance.
Refer to Wccftech