ARM wrestling: how mobile ARM processors crowd out Intel x86 in computers

Not yet on both blades, but desktop processors finally finally got an alternative.

Thanks to which customers will receive decent performance and autonomy in a more budget segment.

Past couple of years interesting for the marketprocessors. Microsoft began to adapt Windows to ARM-processors. Apple is rumored to set out to move its laptops from Intel processors to its own solutions. Mobile Manufacturers of mobile chipsets mastered the 7 nm manufacturing process, and Intel, after a number of postponements of production deadlines, was finally able to master 10 nm.

Intel has been hopeless over the past few yearslagged behind the market, giving primacy in the development of technological processes to companies like Samsung, Qualcomm, Huawei and Apple. At one time, she looked at the smartphone market and was unable to catch up with it either with Atom, having spent $ 10 billion (x86) on the project, or with her own ARM solutions, and disgraced Apple when she did not manage to develop her own 5G modem by this year. Apple reverted to Qualcomm, Intel and the mobile division of its former partner. The power of mobile processors overtook the desktop ones. The desktop market continued to fall, because the user is offered the same thing year after year, only each time more expensive.


</ img>

PC market dynamics over the past 12 years. Source: Gatrner

It is unclear whether Apple has a chance to wrapthis suitcase without a handle is for your own good. But it is absolutely certain that Intel had serious problems in the territory, which the company considered its own for many years. Yes, and the company itself recognized that ARM solutions outperform x86 in some ways.

Hello from the past

While Huawei and Samsung compete in quantityrumors about the planned release of 5-nm processors, Intel since 2014 can not fall below 14 nm. On the 50th anniversary of Moore's Law (in this article we talked in detail about what it is and how it works), the company recognized that it is no longer able to keep up with the designated innovation bar and will develop processors as best it can. While the rest really could, Intel postponed the output of the 10-nm processor for several years in a row, making improvements to the 14-nm chipsets.

This summer, Intel finally introduced the Ice Lake processors on a 10-nm process technology with integrated graphics, which is two times better than the previous one in performance. True laptops on them yet.

Too little. Too late.

Athena Project: Intel Dead Horse

The desktop market has stagnatedbut in the absence of real progress in chipsets, he began to live a strange life. Unable to technological breakthrough, Intel started a marketing epic with ultrabooks. Manufacturers played with materials, improved screens, reduced thickness, but, in fact, could not offer the user anything qualitatively new in terms of characteristics in the top segment. What is and does not change, was offered expensively. Of course, an audience of models for $ 1000- $ 2000 exists, but it is not so large that the segment becomes massive. Intel admitted that with ultrabooks, the company managed to grab only 40% of the laptop market (in this price category) in seven years. The management considers the result to be mediocre and plans to push through another ambitious project in the top segment.


</ img>

In the summer, the company announced the Athena project andshowed on Computex the concept of a laptop of marvelous form, with several screens and a folding case. Devices within the project will cost from $ 1200 and higher, with a permanent network connection, capable of replacing a smartphone, compact, freaky, with AI and other things that make them expensive. Inside - processors no older than Ice Lake. Intel promises partners that it will spend on Athena marketing.

What are the partners? At Computex, companies showed off their laptop options of this kind. But with the other hand they are looking for an alternative. Something that can cheer up the market and bring real sales.

What a twist

In mid-2018, ARM introduced the roadmap of development in the desktop market. The Cortex-A76 core introduced last year surpassed the Intel Core i5-7300U in some tasks. The corporation believes that the experience in mobile devices will help to conquer a new segment, and the Core i5 will yield in performance to mobile solutions this year in Deimos chips (7 nm process technology). In 2020, a transition to 5 nm is planned. Until 2020, the company promises a 15% increase in productivity.


</ img>


</ img>

ARM Processor Development Roadmap

Microsoft, ARM, Qualcomm: the hearts of three

Microsoft, like Intel, also didn’t work outbusiness with the mobile direction. It turned out that the giant operating systems simply does not know how to make convenient interfaces, and even subsidies and the Nokia brand could not sell smartphones on Windows. The project was closed, pouring billions into the pipe. The company relied on tablets on stripped-down Windows (which didn't know anything) and fast and cool, but insanely expensive ultrabooks. We have already talked about them above. Small market segment and lack of portability again. While smartphones were increasing screens and days of work, the majority of Windows solutions could not reach the evening, and the claimed autonomy hours did not correspond to reality. Not only the hands of Microsoft engineers and gluttonous Chrome are to blame for this, but also Intel's long-standing partner with x86 solutions.

In 2017, Microsoft began to adapt Windowsfor ARM architecture. Qualcomm, in turn, introduced the chipsets at the end of 2017, partners - HP, ASUS, Lenovo, Samsung and others - the first laptop models.


</ img>

At the end of last year, Qualcomm introduced a chipsetSnapdragon 8cx. Apparently, this is exactly the same Deimos with eight Kryo 495 cores (four high-performance Cortex-A76 and four energy-efficient Cortex-A55). In benchmarks, it is comparable to the Intel i5 8250U. That is, with a modern laptop of a higher than average level, ARM laptops will be able to give the user the best experience in autonomy (real days of work when connected to a mobile network) and comparable in performance in the category up to $ 1000. Compare this to the Athena project, which is about the lower bar of $ 1,200.

Of course, ARM solutions will not work for everyone.to users. But for Microsoft, which was losing users in favor of iOS and Android tablets, for vendors who found it difficult to drive sales of laptops, especially in the middle segment, for users who wanted to progress but did not get it - for everyone this is a new interesting round of market development .

Apple: Brutus and you

This year, the find Bloomberg somehow hushed up, inIn any case, in the media space, but in the past, it boomed in many non-industry media. According to rumors, Apple plans to transfer the MacBook to mobile ARM processors next year. This will be done (if any) in order to reduce dependence on Intel, bring the ecosystem closer and simplify porting applications between devices.

This year, introducing the new iPhone and chipset,Apple said this is the most powerful solution in the world of mobile devices. Apple loves bravado and tests on the Internet show that you can’t talk about absolute superiority, at least Samsung, Qualcomm and Huawei have alternatives of the same level. But in this case, the trend is more important and the fact that the performance of the iPhone chipset is also very high. IPad performance is enough to mount 4K video. Samsung's flagship tablets and smartphones paired with a Dex station (you can read our Galaxy Tab S6 and Galaxy Note 10+ reviews) are already capable of replacing a computer for most tasks. Huawei will probably also work in this direction if the sanctions are lifted.

In the dry residue

Special project

Test: true or myth? Test your smartphone payment knowledge

There are no open position reports in open access.the market of various manufacturers of mobile chipsets, and there is no single report on the market of mobile and desktop solutions in principle. But it seems that the future of x86 is not as bright as Intel seemed when the company decided to enter the smartphone market with Atom in full confidence in the invincibility of its development. Intel is unlikely to die in the coming years. In the end, there are other niches where the company's products are in demand, and there are enough demanding or conservative users. But it seems that the market for desktop solutions is waiting for changes and revival, and we are more interesting and inexpensive laptops on ARM. Probably immediately with 5G.