Yesterday’s announcement came almost exactly 15 years after Steve Jobs and Paul Otellini, a former president and CEO of Intel, announced the collaboration of the two companies on a stage, patting each other’s shoulders. The almost nostalgic recordings of the time also provide a clear answer to the reasons: Jobs and his team no longer saw a perspective on PowerPC chips, which slowly began to undermine the success of Macs. Meanwhile, Intel was just beginning its unprecedented march, with the release of the first Core processors a few months later, bringing the market to a new level in both computing power and efficiency. Looking back over the past decade and a half, it is indisputable that Jobs and his company have just wandered to Intel processors and its x86 instruction architecture at the best moment.
However, consumer computing has undergone a significant transformation since Macek's previous 15 years. Apple introduced the iPhone in 2007, which, in the absence of a suitable alternative, did not land on an Intel processor, but on an ARM-based development purchased from Samsung. At the time, it was presumably not even in the most terrible nightmare of the chip maker in Santa Clara that it would simply be incapable of finding a catch in the smartphone market, which swelled enormously in just a few years. But that is exactly what happened in the end. The atomic outing brought the biggest failure in Intel’s history to date, with nearly $ 20 billion in money thrown out the window. Being a huge, capital-intensive company, the chip maker did not give up. With his modem upgrades, he tried to sneak back through the back door into the pockets of shoppers. Based on last year’s developments, this attempt was not crowned with much greater success either, with only Apple ultimately willing to buy from radio chips. The end of the modem adventure is also well known, with Apple acquiring a majority stake in the business in 2019, making Intel virtually off the shoreline of the smartphone market, with a good chance forever.
While Intel has been desperate for the past roughly 10 years. tried to find a catch in the mobile market, until then the iPhone and with it Apple rose to astonishing heights. Thanks to smartphones, billions of dollars have flowed regularly into Cupertino, which is no better illustrated by the fact that iPhone traffic was rubbing as much as $ 40 billion in prominent quarters. Thousands of money have been wisely invested by Apple, which in recent years has begun to build its hardware design team with great determination. The jewel of this is the processor development department, which has set a new standard for the complete smartphone market over the past roughly 5 years. In addition, the design team had to focus solely on the company's own needs, as Apple does not sell “A” -chipped chips as a stand-alone component, and there was never a risk that a competing manufacturer would pack a more powerful application processor under iOS. .
However, the lack of direct market pressure had no effect on the work of the design team, which was able to increase CPU computing performance by 20-30 percent from year to year, and in extreme cases by up to 50-60 percent, with no change in dissipation. considerably. With this, Apple was the first to break the long-standing stereotype that the performance of ARM processors is inferior to x86 chips at the cost of low dissipation and chip size.
in part, this may consist of, while self-designed processors may also have the added benefit. Thanks to the customized chips, the dependence on quasi is completely eliminated. Until now, Macek's roadmap has always had to be aligned with Intel's (formerly IBM's or Motorola's), which is likely to have long pushed Apple's beak, which has been slowly but surely vertically integrating over the past 10 years. Taking advantage of independence, Cupertinos could respond more flexibly and even faster to the current needs of the market, as we have seen several times with the iPhone and iPad. An additional benefit of self-designed chips may be that they can optimally outperform Intel processors, a differentiating factor that may have been another compelling argument for switching.
The platform underlying the current Macs has been sourced from almost any other PC maker. you can buy a similar, sometimes even more attractive garnish, often cheaper. With Apple's proprietary processors, the basic hardware similarity and interoperability will be eliminated, and if the company's CPUs designed for notebooks and desktops are to stand out, it could be another strong reason for Macs. (Of course, the opposite is not to be missed, ie that in some segments Apple's chip may even lag behind Intel's current development.) Although the Cupertino company has not yet talked about specific plans, there is a good chance that more dedicated, fixed-function accelerators will be available in ARM.
Another, undisguised goal of the company is the three big platforms, namely the iPhone and the iPad, Macs, which are a kind of guarantee that the operating system and development tools are made by the same Apple as the processors. and the convergence of the Mac, one important step of which is the common, ARM instruction architecture. An earlier prelude to this was Project Catalyst (formerly known as Marzipan), which gave iOS / iPadOS and macOS a common development platform. With the latter, Apple wants to make the desktop operating system more accessible to iOS developers. This is because while an incredible number of developers build better applications for the mobile platform, the knowledge gained in this way is difficult to transfer to the world of Macs. Although many things (such as Objective-C and the Swift language) are similar between the two platforms, and iOS has inherited a lot from macOS at the API level, macOS is quite foreign to the one-time mobile developer.
Thanks to Catalyst and the same CPU instruction architecture, software developers will have to put together a single application that works on a touch screen or a mouse-keyboard, depending on whether it's running on an iPhone, iPad, or Mac. This is a significant change from the previous situation where two different applications need to be written for iOS and macOS, respectively. Easy interoperability can optimally result in a true win-win situation: developers can reach more users without investing additional resources, while the number of applications available for macOS jumps significantly, presumably to the delight of users.
Last but not least, In many cases, the processor, which is in many cases the most expensive part of notebooks, will not have to be paid for by an external supplier, in this case Intel, as the design takes place in-house, so there are only license fees and chip production costs. 19659004] Well-proven choreography with some question marks
Based on yesterday's presentation, Apple pulled the well-proven PowerPC-> Intel migration scenario out of the account. According to Craig Federighi, who is responsible for the company's software development, the full migration will take about 2 years, or roughly 2022, during which time various Macs based on in-house processors and native ARM applications will gradually appear. Almost nothing specific has been revealed about the former, ie the products, but it is certain that the first Mac with an Apple processor will be on the market this year. To some surprise, Tim Cook concluded that audiences can expect a new Intel-based Mac (or Macs?), But the identity of these products is still unclear for now.
For many, the question of what will happen with support for x86 Macs is now more pressing. . Apple promises that machines with intelligent and proprietary processors will receive newer macOS versions in parallel, at least for a while. However, the company has not yet said anything specific, i.e. it is not known whether the aid is for 2, 3 or, for example, 5 years. It is worrying that Steve Jobs also talked about 2 years of the PowerPC-> Intel migration, which was finally put into practice: the smart switch was announced in June 2005, and the last MacOS 10.5, which also supports PowerPC, was released in October 2007. and 10.6 “Snow Leopard” was only installed on smart machines.
Running x86 applications is also an important issue. Apple thought of this, as Rosetta, familiar to many from the previous migration, is reviving with a second version, so in theory, there will be no obstacle to ARM Macs running x86 applications. However, support is still in question in this case, as the first version of the Rosetta was shot by Apple in 2011 after roughly 5 years, so chances are x86 compatibility won't last forever either. It is not yet clear whether, like other players (such as Microsoft), Apple will disclose how long it will provide support. In the absence of this, it is currently not worth thinking about a smart Mac for more than a few years.
It may also be uncertain that Apple will focus on the ARM version of software, unspoken, as developers begin, which developers will certainly follow over time. . According to Craig Federighi, this has already begun, as Apple's macOS applications run natively on ARM one by one. However, according to Federighi, this only requires the latest Xcode, which can create both x86 and ARM binaries for the application, so for most developers, the transition can take just a few days of extra work. Microsoft and Adobe have already begun the transition, which will make the most popular applications available in native form as soon as the first ARM Macs are released. It is also good news that virtualization does not have to be abandoned, so with Parallels, even Linux and, in theory, even Windows 10 ARM can be run on next-generation Macs.
The first Apple's processor product is expected to be another MacBook Air, maybe a MacBook Pro 13, which in terms of processor is likely to share with the central unit of the next iPad Pro. Accordingly, there is a good chance that the A14X, or a parameterized variant of it, will be included in the first ARM notebooks. Based on the A12Z released in the development kit and the possibilities of the upcoming 5-nanometer manufacturing technology, this could mean a processor of up to 12-16 cores, and there will probably be no shortage of executives and computing power. For now, the integrated GPU can only be guessed at, as is Apple's plans for GPU supplier AMD in the absence of its own chips.
A more interesting question is that in a desktop environment, i.e. the iMac (Pro) and What kind of processor can a Mac Pro get after Intel's multi-core and sometimes high-clock products. This, i.e. the desktop PC, workstation area, is still unknown to Apple's chip design department. Is it worth designing a dedicated, custom processor specifically for larger desktops, and can the hundreds of millions of dollars in investment be recouped from sales of iMac (Pro) and Mac Pro products alone? If not, Apple may even follow the example of AMD Zen, which uses a single chiplet and a stand-alone, less expensive I / O chip to solve the growing burning problem of design flexibility and manufacturing economy.
Intel: our product is better
An interesting question is how Apple's move will affect Intel's position in the longer term. The company is still able to absorb a few billion dollars a year in lost traffic, but any longer-term damage could be a bigger headache for the processor manufacturer. ARM CPUs, which dominate the smartphone market, have so far been unable to gain a foothold in Intel's highest-revenue PC market, and while a joint attempt by the Qualcomm-Microsoft duo has sparked a few tiny rays of hope, no success is yet possible. If, on the other hand, Apple were able to create something astonishing in a market that has long been free of serious excitement, it could even trigger an avalanche, prompting other manufacturers to respond. It's worth recalling that the Windows-based competition has long lagged behind the once-seemingly astonishing 8-10-hour battery life of MacBooks, which also build on the Intel platform.
processors. Nevertheless, it now seems that Intel would have regretted less if Apple had switched to AMD, so that the x86 ecosystem would not have been damaged or, in parallel, ARM would not have been able to hold on to the PC market. In a brief response, Intel noted that it will support Apple and its products until the end of the contract (which date is not public), while believing that Tiger Lake, which will be available by the end of this year, will provide the best experience and value on the market for all users. , and for developers.
Gellert is Technology Editor at Counting News Media and contributor at other major tech publications. Her interests includes testing new gadgets and reading.