By stopping the development of low-end Atom processors, is Intel finally giving up on the mobile market?

Recently, there was news that Intel will completely cancel the development of the Broxton and SoFIA Atom processor product lines. Yesterday, Intel officially confirmed the news, and these two chips have become the first victims of Intel's business transformation.

▌ Broxton and SoFIA

Broxton and SoFIA are both chips developed for mobile terminals, and both were officially announced by Intel in December 2013. The former is for high-end mobile products, and the latter is for low-end mobile products:

• Broxton was originally planned to be launched in mid-2015 using the new Goldmont architecture and 14nm process, but its shipment has been delayed until now.

  
  

• SoFIA is Intel's first mobile SoC with an integrated baseband chip. The first generation product equipped with a 3G baseband was launched in the second half of 2014. However, the SoC with a 4G baseband originally planned to be launched in 2015 has also been postponed to date.

Intel mobile chip development roadmap

As benchmark products in Intel's mobile chips, Broxton and SoFIA represent the future of Intel's mobile chip business. In addition, an Intel spokesperson said:

"Resources used to develop the Broxton and Sofia chips will be redirected to products that deliver higher returns and advance our strategy."

Apparently, after evaluating the mobile market and its own products and performance, Intel has decided that it is time to gradually abandon the mobile chip market. However, this also means that Intel's previous investment of billions of dollars in mobile chips will go down the drain.

So why is Intel so determined to give up this market that it has been cultivating for 8 years?

▌ It's a pity to eat it and throw it away

"It's mainly a matter of ecosystem. Putting Intel chips in a mobile phone is as difficult as putting an ARM chip in a desktop PC."

In this regard, Tie Liu, a columnist for Leiphone.com (search for "Leiphone.com" official account and follow it), who has in-depth research on the chip field, expressed his views.

"In the mobile terminal market, few apps or APIs (application programming interfaces) support X86 architecture chips, and all content providers are developing around the ARM architecture. Although Intel may still have hope if it continues to invest in the tablet industry, in the long run, I am still not optimistic about Intel's development prospects in this market."

If we only look at the current smartphone market, apart from ASUS, the only company that still sticks with Intel chips (it has now abandoned them, with news showing that Zenfone 3 will use Qualcomm Snapdragon 650 SoC), other companies have already switched to the ARM camp.

The reason for this is that, in addition to the above-mentioned ecological reasons, the problem of excessive power consumption caused by overperformance/no support for baseband-bundled SoC is also a major reason. (The only SoC bundled with baseband still uses 3G chips, and is produced by OEM)

Zenfone 2 may be the last "Intel inside" phone in the world

Even with a huge subsidy of $1 billion, the tablet market, which shipped 40 million tablets in 2014, has seen no improvement in mobile chip products due to the pressure from two-in-one notebooks and ultrabooks in recent years.

"Intel's strengths lie in desktop PCs and servers. After all, the smartphone market has reached its ceiling, while the server and smart wearable markets are continuing to grow with the explosion of big data, cloud computing and the Internet of Things, so it is more important for Intel to do well in these two markets."

After the financial report was released, Intel CEO Brian Krzanich announced that Intel would undergo a large-scale business transformation, shifting its main business from PC platforms to cloud computing platforms and Internet of Things platforms. The fate of the mobile chip business, which was not mentioned at all, is predictable.

"In addition to its performance advantage in the server market, another major advantage of X86 is its ecosystem. Currently, most data centers use the x86+linux ecosystem. Since ARM 32-bit instructions and 64-bit instruction sets are incompatible, ARM basically has no market. However, the situation in the mobile market is completely the opposite."

▌ Although Mobile computing chips have been abandoned, but the mobile communication chip business is still retained

So has Intel really given up on the mobile market? Brian Krzanich's words in his blog a few days ago may provide an explanation:

Intel will not give up (the mobile market) and will focus on how to guide the entire industry to 5G. This may be a huge challenge, but it will also be an opportunity for Intel to finally leave its mark in the mobile market.

In fact, strictly speaking, they are no longer in the same market. Because the Atom processors that Intel made before are computing chips, and its competitors are SoC chips such as Qualcomm Snapdragon and MediaTek; however, if it is to build 5G chips, it will challenge the baseband communication chips of companies such as Huawei and ZTE. So, can Intel achieve its wish to overtake in 5G technology?

"It's very difficult, just as difficult as Loongson trying to make a comeback in the PC market. After all, Intel is not a major communications company. In terms of baseband, it is very likely that it will not be able to compete with Huawei, ZTE, Qualcomm and other old players in the baseband field. Moreover, companies like Huawei and ZTE are directly involved in the formulation of 5G standards. Manufacturers that do not participate in the formulation of standards will have to pay expensive patent fees if they want to get involved."

Although Intel's reputation in the field of mobile communication basebands is not as loud as that of x86 chips, Intel has also launched several basebands since acquiring the veteran wireless manufacturer Infineon, such as the third-generation five-mode baseband XMM 7360 at MWC 2015 and the XMM 7480 launched in February this year.

Earlier, there was also news that in order to curb Qualcomm's baseband monopoly on the iPhone, Apple decided to use Intel's baseband chip XMM 7360 on 30-40% of the iPhone 7. If Intel can take advantage of this opportunity to win more baseband orders from Apple, then Intel still has hope of keeping a foothold in the mobile market.

▌ Recommended articles:

How are the "soft costs" of chips composed?