Opinion: Alphabet wants to make its own smartphone chips — what in the world is the company thinking?

Alphabet said Monday it will take its silicon building in-house when it launches a new flagship line of Google Pixel phones.

The company’s move creates as many questions as it does answers. Perhaps the first being, with only about 2% of the U.S. smartphone market and no real sign of that changing, does a bet like this make sense?

See also: Microsoft, Google and other U.S. companies requiring workers to get vaccinated before returning to the office

While much of the analysis of this news pointed to it as a positive step for Alphabet
I interpreted it as a desperate attempt for the company to make headlines about its poor-performing line of mobile devices and to do so by taking a risk that even Apple
isn’t yet willing to take when it comes to its vertical integration into semiconductors. 

Apple’s foreshadowing

Apple made headlines over the past few years with its departure from Intel
and the introduction of its M1 architecture. This transition from Intel to homegrown processors built on the increasingly popular Arm architecture was done methodically over several years to ensure that the change would go smoothly. For the most part, the M1 launched without significant incident, but its critics would be the first to point out that the transition was anything but flawless. 

See also  IPO market faces busiest week of the year — if all 19 deals on calendar are pushed through

It’s also worth pointing out that Apple, while it has incorporated a number of its cores to handle capabilities like image sensing and graphics, has left the modem and radio, oft-referred to as RFFE, to Qualcomm
due to the complexity of building a modem-RF system for 5G, which is incredibly complicated and proper function is imperative to get devices certified by different carriers like Verizon
and others. 

For some reason that I simply cannot understand, Alphabet wants to abandon not only the Qualcomm Snapdragon, which most simply can be considered the full set of systems to power all the phones features, and ditch it for its own set of cores, but allegedly also abandon Qualcomm’s proven modem-RF system to use (rumored) Samsung’s Exynos, which has almost zero market penetration and has run into issues with power management and carrier certification. Even Samsung uses Qualcomm’s modem-RF system for large portions of its flagship devices, including its U.S. flagship devices. Apple does as well. 

Returning us to the question: What in the world is Alphabet doing here, and does it make any sense? 

Vertical integration

The short answer is yes, and no, but mostly no. At least for Alphabet. 

See also  Bitcoin at $250,000 in a year? This 'rocket fuel' will help it get there, says Goldman Sachs alum

The argument to favor this change comes down to profitability and control for Alphabet. Apple has proven that deeper vertical integration can be lucrative, and its Bionic chips have continued to improve with each generation. Alphabet, which has built a very limited set of custom ASICs (application specific integrated circuits) for artificial intelligence (AI), wants the market to see it as every bit as capable of this type of move as Apple. 

Furthermore, we have seen Amazon’s
AWS and Alibaba
build Arm-based solutions for their cloud offerings, and Microsoft
has also been touted to be following suit. For the data center, this seems to make sense, with AWS, for instance, seeing success from its Graviton line of CPUs and Trainium series of ML chips. No one would be surprised to see Alphabet significantly increase its efforts to launch or further enhance custom silicon development for its cloud and AI offerings. But servers, and even notebooks, aren’t mobile devices. 

To launch the M1, Apple spent years on R&D and made several key acquisitions, including Dialog Semiconductor in 2018 for $600 million to add to its internal chipmaking and systems capabilities. It even acquired the remnants of Intel’s mobile modem business upon deciding to move away from Intel on the horizon of 5G. But, with all of that IP, and investment in tow, and even a multi-year legal feud with Qualcomm, Apple still recognized there were some things better left to be done by Qualcomm. 

See also  https://www.marketwatch.com/story/june-2021-report-on-global-nails-and-screws-market-statistics-cagr-outlook-and-covid-19-impact-2021---2023-2021-06-15

Perhaps Alphabet figures with almost no market share and no notoriety around its Pixel smartphone lineup that it doesn’t matter if the Pixel 6 series is a disaster.

And perhaps Google will surprise the world by using its cores plus a lesser-quality modem-RF system to somehow put out a device that will outperform Apple, Samsung and dozens of other handset makers that have entrusted Qualcomm to provide the guts of their devices.

I doubt it, though, and if nearly a decade of insignificant market penetration with their mobile devices serves as an indicator of how well this move will go for Alphabet, I think it’s safe to chalk this up in the L column.

Daniel Newman is the principal analyst at Futurum Research, which provides or has provided research, analysis, advising, and/or consulting to Nvidia, Qualcomm, Microsoft, Amazon and dozens of other companies in the tech and digital industries. Neither he nor his firm holds any equity positions with any companies cited. Follow him on Twitter @danielnewmanUV.

View more information: https://www.marketwatch.com/story/alphabet-wants-to-make-its-own-smartphone-chips-what-in-the-world-is-the-company-thinking-11628003771

Articles in category: moneyist

Leave a Reply

Back to top button