Intel recently abandoned its plans to produce and sell 5G modems for smartphones after its main customer, Apple, announced on April 16 that it would start using Qualcomm modems again. Apple has used the company's modems in the past, but switched to Intel products solely due to litigation with Qualcomm over patents and high licensing fees. However, Intel's achievements in 5G are significantly inferior to its competitor, and Apple does not want to waste time and lag behind Android manufacturers due to the unpreparedness of its partner to master the new technology.
Qualcomm has already released its first 5G modems, while Intel planned to start production of the first copies only in 2020, which, if the partnership between Intel and Apple continued, could lead to the introduction of the 5G iPhone about a year after the first Android devices supporting the new standard connections. To make matters worse, analysts at UBS and Cowen have warned that 2020 could be a fairly optimistic forecast for Intel that doesn't match reality at all.
Intel disagreed with UBS and Cowen's forecasts, but Apple's decision, which clearly prioritizes the early release of the new iPhone over winning the legal battles with Qualcomm, indicates that analysts were probably not far from the truth. The situation can be considered the second failure of Intel in its attempts to enter the mobile device market. Let's take a look at Intel's past failures and what they could mean for its future.
How Intel lost its chance in the mobile market
More than a decade ago, Intel said that Apple would not be able to sell significant enough volumes of the iPhone, and therefore refused to produce processors for its first smartphone. Apple eventually ordered processors from Samsung before developing its own A-series processors, which ended up being made by both Samsung and TSMC.
Intel then neglected the meteoric rise of ARM, which licensed low-power chips to mobile chip makers like Qualcomm. In fact, at one time Intel had its own microarchitecture for ARM processors - XScale, but in 2006 it sold it to Marvell Technology. At that time, Intel decided it could use its leadership position in the PC and server markets, which mostly use the x86 architecture instead of ARM, to promote its Atom x86 processors for mobile devices.
Unfortunately, Intel x86 processors were not as power efficient as ARM processors, and mobile device manufacturers preferred battery life over performance advantages. As a result, customers turned to ARM chip makers such as Qualcomm and Samsung. Qualcomm soon integrated the modem and graphics core into an ARM chip in its Snapdragon processor family, which became a cost-effective all-in-one solution for most smartphone manufacturers. By the beginning of the new decade, ARM processors were used in 95% of all smartphones in the world, and Qualcomm became the largest manufacturer of mobile chips.
Instead of giving up, Intel tried to get back into the smartphone market by subsidizing hardware manufacturers that used Atom chips. Within three years, about $10 billion was spent on subsidies to occupy no more than 1% of the market. When Intel cut subsidies, OEMs, as expected, went back to ARM chips.
In mid-2016, Intel finally discontinued the Atom SoC for smartphones. In the same year, the company began shipping 4G modems to Apple, which distributed orders between Intel and Qualcomm. However, Intel's modems were noticeably slower than Qualcomm's, forcing Apple to limit the speed of the latter to eliminate differences between its own phones.
Therefore, it is not surprising that with the already emerging lag, Intel lost in the 5G race as well. The company clearly couldn't match Qualcomm's experience in this area, and Intel's constant problems with underproduction of chips on the 14nm process, which is used, among other things, by its own modems, only exacerbated the problem.
What does this failure mean for Intel?
Apple's decision not to partner with Intel comes as no surprise, but Intel's confidence in its path raises questions about the company's management.
On the other hand, Apple's decision could help Intel improve its 14nm shortage. Also, the loss of Apple as a customer for the company's future 5G modems should not noticeably affect its revenues, which are primarily focused on the PC market (52% of Intel's revenues in 2018), especially since production has not even begun yet. It could also cut spending on research and development, which used up nearly a fifth of Intel's revenue last year, and allow Intel to spend more money on promising technologies where the company hasn't lost the fight yet, such as self-driving cars.
Interestingly, shareholders and the market seem to be thinking in the same direction, given that the decision to stop supplying 5G modems caused a slight increase in Intel shares, instead of a seemingly expected fall, as analysts believe that this will allow the company to reduce unnecessary costs that reduce its net profitability.
Intel is not completely abandoning the development and supply of modems. The company is still planning to release 4G and 5G chips for PCs and IoT-enabled devices. However, the loss of Apple's orders was the company's second setback as it failed to gain a foothold in the huge smartphone market. Let's hope that Intel has learned its lesson and will focus more on innovation without betting on its default superiority of the competition, as it did with Atom.
Source: 3dnews.ru