[ad_1]
Before Apple’s Double 11 conference, the Mac was the only product line that didn’t use its own processor. Chip replacement is not an easy task, it not only involves the transfer of the underlying architecture and the rewriting of the system, but also the compatibility of a large number of third-party software.
In 2005, when Apple left Motorola and switched to Intel from PowerPC, it experienced monotony.
At that time, the entire migration time took Apple 4 years. Starting with the first generation MacBook Pro, it wasn’t until the first OS X Snow Leopard system that it didn’t support the PowerPC architecture. Apple completed the complete transfer of Mac hardware and software to the X86 architecture.
Apple is willing to take risks, but it simply wants the Mac to ditch the old technology and get more competitive chips to improve performance.
▲ The purpose of Apple’s use of self-developed chips and transfer architecture this time: to obtain powerful performance, personalized technology and better power consumption management.
Now Apple wants to leave Intel and adopt self-developed chips for the same reason. It’s just that Apple’s current situation, simply pursuing performance, is obviously not the sole purpose of self-developed chips.
What are the characteristics of the Apple M1 chip?
The M1 chip is Apple’s first chip designed specifically for Mac, and it is also the first chip based on the ARM architecture on the Mac platform.
(Source: Apple, same below)
It is no longer a pure CPU processor, but instead uses SoC packaging technology like the iPhone and iPad A-series chips, and the common Snapdragon chips found in Android phones.
This means that the CPU, GPU, memory and other modules are integrated for a higher degree of integration, which is also beneficial for freeing up internal space in the computer.
Apple has revealed some details of the chip for the first time. First of all, in terms of manufacturing process, M1, like A14, is based on the latest 5nm process, but compared to the 6-core design of A14, M1 uses “4 + 4” cores to form a CPU 8-core and an 8-core GPU, up to 16 billion transistors. , Which is also 35% more than A14.
Just looking at the architecture, Apple’s M1 is actually very similar to earlier derived A-series chips such as the A12X / A12Z previously used in the iPad Pro and the iPhone A12. It also improves performance by increasing the amount of CPU / GPU cores.
From this point of view, M1 and A14 have many connections.
But since it is used on the Mac platform chip, M1 will obviously also get better heat dissipation conditions than A14, along with the change in the size of the first and second caches, and the use of high bandwidth, often indicates a better performance version.
At the press conference, Apple emphasized the “performance per watt” advantage of the M1 chip, and also compared it to conventional notebook processors in the industry. Among them, with a power consumption of less than 10 W, the M1’s CPU performance is twice that of competitive PC processors, and M1 only needs a quarter of the power consumption to match the performance maximum of competitive PC processors.
The GPU situation is similar, the same performance is twice as bad as the 10W power consumption and the same level of performance is achieved with lower power consumption.
However, there are still doubts. For example, the size of the M1 chip and the core clock are still unknown, Apple’s competitor products are also very vague, and if different heat dissipation affects the performance of the Air and Pro molds, these should be evaluated and be answered later.
In addition to the performance improvement, Apple also specifically mentioned the “unified memory design” of the M1 chip.
In traditional PCs, the CPU and GPU memory are physically independent of each other, resulting in some shared data that must be exchanged and allocated once between the two memory pools.
But this time, the M1 chip adopts a “unified memory design” similar to other Apple devices like the iPhone. In fact, the CPU and GPU share the same memory pool, which not only makes data exchange more efficient, reduces latency, but also saves power. .
The 16-core neural network engine is similar to the A14 architecture. In recent generations of iOS systems, many features are inseparable from machine learning, such as image synthesis, gallery sorting, and even power management. There are similar technologies behind them.
Today, the value of porting machine learning to Mac is also the same. With the hardware-level support of the M1 chip, such as improving the speed of computer photo synthesis and improving the accuracy of speech recognition, it can be used by Mac. It is ported from various iOS platforms. The application, machine learning technology can also be used on the Mac platform.
The remaining features are also very simple, nothing but support for USB 4, PCI 4.0, NVMe storage and other features, as well as Apple’s special encoding and decoding engine and security isolation area technology, in fact, many of them have appeared on Apple’s T2 chip.
The dedicated codec engine will significantly improve the running efficiency of Apple’s own image editing applications, such as Final Cut Pro, and the increase is also multi-level; the safe zone is to apply the privacy protection that Apple has always attached to Mac.
What has changed in the first batch of Macs using Apple chips?
Compared to the high-profile M1 chip, the first three Mac products that use the chip aren’t too dazzling. The problem keeps popping up. Many people originally expected that the architecture change and the low power features of the ARM architecture would make Mac products a major change.
But at this point, obviously, Apple is not ready to adjust the appearance of the MacBook. This time, the three new Macs basically use the old molds, which can be seen as a “change the chip without changing the case” upgrade.
▲ Upgrading the M1 chip to Mac is primarily reflected in performance and battery life.
There are still bright spots. After all, ordinary users do not care too much about the multiple performance improvements, but if the battery life shoots up, the screen responds faster and is directly compatible with the iOS application ecosystem, they are actually obvious and practical changes.
To emphasize the improved battery life offered by the M1 chip, Apple’s presentation will directly use the metaphor of “longest battery life ever” to highlight the longer battery life of the MacBook Air and MacBook Pro. 13 inches.
Under normal circumstances, the MacBook battery life is about 10-12 hours. If it is used a lot, it can die after 6 to 8 hours. The M1-chip MacBook Air now has a battery life of up to 18 hours, and the 13-inch MacBook Pro has shot up to 20 hours.
If the data is true, you can get a new Mac in the future, just type and surf the web, and you won’t need to charge anything for a day off. The habit of using laptops in many settings will change as well.
The arrival of the M1 chip also gives the MacBook the same quick wake-up capabilities as mobile phones and tablets.
Craig Federighi, the vice president in charge of software at the presentation, showed a clip: He reached out and turned on the MacBook, the screen instantly lit up, and the color of the tablecloth was directly on his face, emphasizing the “quick response” feature.
All-weather battery life, instant wake-up, and excellent heat dissipation are also features that PCs with ARM-based chips have liked to show off in the past. For example, the Qualcomm Snapdragon 8cx laptop and Microsoft Surface Pro X weren’t new.
But for MacBook, it is the first time the ARM chip is used; It is also the first time that these characteristics have been obtained.
As for the Mac mini, as a desktop system, the M1 chip upgrade naturally does not imply endurance, but rather focuses on improved performance and energy savings.
Of course, Apple also lowered the starting price of the new Mac mini, which is no small surprise. If you’re interested in experiencing the M1 chip, Mac mini will be a good entry-level product.
We also noticed that the M1 chip has changed the Mac software ecosystem. In the last two years, although Apple has worked hard to organize a launcher to bring iOS applications to the macOS system, it is not easy because the underlying hardware of mobile phones, tablets and computers is still different.
▲ iPhone and iPad applications can be run directly on the Mac equipped with the M1 chip, but the experience must be tested.
After Mac appeared on the M1 chip, it directly solved the application interoperability problem caused by hardware differences. Apple has stated that all Macs equipped with the M1 chip can run applications directly on iOS.
Specifically, how to match, according to Apple, the swipe, click and tap gestures will be implemented on the Mac with a trackpad and keyboard. It seems like it won’t be difficult to team up with a Mac to play “Glory of the King” in the future.
In general, the first batch of products equipped with M1 chips focused on the interior rather than appearance. In fact, Apple is also very cautious about its “product selection.” For example, Air and mini are positioned at the mass consumer level. The professional product line only replaces the 13-inch MacBook Pro with an M1 chip, which does not affect high-end products.
Considering that professional users depend on specific software and workflows, they will have higher requirements for the device runtime environment. Apple will not change the environment hastily until the architecture transfer is complete and the software is fully compliant.
But it is foreseeable that starting next year, the 16-inch MacBook Pro and even iMac product lines will definitely launch Apple’s self-developed chip version. It’s only a matter of time before the entire Mac series abandons the Intel and x86 architecture.
The name M1 also indicates that in the future, Apple will continue to introduce chips like M2 and M3, improve the Mac experience, and allow Mac to test new hardware that couldn’t be done before. These are the possibilities of the M series chips.
As long as this improvement gets the Mac rave reviews, even if it ditches x86, it won’t cut Intel’s sales too much, but it will be a huge shock to other manufacturers.
More than 40 years ago, Apple began manufacturing personal computers and gradually morphed into a company that relied on the iPhone and various services. Many people have felt that personal computers account for less than one-seventh of Apple’s quarterly earnings, and there is no need for reform. But now, relying on self-developed chips, Mac will also embark on the same upgrade pace as the iPhone and iPad, and will have the opportunity to reshape the original software experience under the unified and integrated ecology.
At the “One More Thing” conference, Apple once again demonstrated to the industry its ambition to make chips, Apple also has new answers for the future of Mac.
(This article is authorized for reprint by Ai Faner; the source of the first image: Apple)
Other readings:
[ad_2]
