Google (NASDAQ:GOOG) (NASDAQ:GOOGL) Glass will get a major hardware upgrade soon, according to The Wall Street Journal. Google will reportedly replace its original Texas Instruments (NASDAQ:TXN) processor, licensed from ARM Holdings (NASDAQ:ARMH), with one from Intel (NASDAQ:INTC) next year.
That swap was stunning for two reasons. It's proof that Intel -- which was marginalized by ARM's disruptive licensing model in smartphones and tablets -- is becoming a major player in wearable devices. It also demonstrates that Intel's new chips are just as power efficient as ARM's designs. Let's see what this decision means for Intel, ARM, and Google.
Why Google Glass matters to Intel
Intel lost the smartphone and tablet markets because ARM designs were widely considered to be cheaper and more power efficient. Although Intel's Atom chips improved over time, major mobile players like Apple, Samsung, Qualcomm, and Nvidia had all hopped aboard the ARM bandwagon.
To fight back, Intel subsidized Atom-powered smartphones, tablets, and netbooks by providing hardware partners with discounts on Atom chips, co-marketing agreements, and financial assistance in redesigning logic boards for Atom chips. This costly strategy caused its mobile chip division to incur $3 billion in losses in 2013 and will likely cause an additional $4 billion in losses by the end of 2014, according to Morgan Stanley analyst Joseph Moore.
Over the past year, Intel invested heavily in wearable devices. It acquired health tracker maker Basis Science in March, which led to the recent launch of the Basis Peak smartwatch. In August, it partnered with SMS Audio to launch biometric earbuds. The following month, it partnered with Fossil to make smartwatches and unveiled a luxury bracelet/smartwatch (adorned with pearls, obsidian, and snakeskin) co-developed with Opening Ceremony. Intel also recently signed a deal with Luxottica to develop fashionable smart glasses.
Google's decision to install an Intel chip in Glass is a confirmation that Intel has a foothold in the wearables market, which Grand View Research projects will grow from 2 million shipments in 2013 to 135 million by 2020. However, it's still unclear which Intel chip -- the Atom, the tiny Quark, the cellular/processor hybrid SoFIA, or something entirely new -- will be installed in Glass.
Why Google Glass matters to ARM
Google Glass currently uses Texas Instruments' OMAP 4430, an ARM Cortex-A9 chip.
ARM's popular 32-bit Cortex-A designs are commonly used in mid to high-range devices. Cortex-A chips accounted for 18% of ARM-licensed shipments last quarter while 53% came from its lower-end ARM7/9/11 designs, which are commonly installed in lower-end devices. The newest top-tier mobile devices, like Apple's iPhone 6, use ARM's 64-bit ARMv8-A designs.
ARM's business is split into licensing fees (upfront payments) and royalties (from each device sold), which respectively accounted for 45% and 46% of its revenue last quarter. Over the first nine months of 2014, ARM's licensing fees rose 31% year-over-year, but royalty fees only climbed 4%. This means that more companies are signing up with ARM, but fewer devices are being shipped. To exacerbate that problem, more lower royalty ARM7/9/11 chips (1% to 1.5% per unit) than higher royalty Cortex-A chips (1.5% to 2%) were shipped. ARM hopes that more customers will upgrade their devices to 64-bit ARMv8-A designs, which pay a higher royalty rate exceeding 2% per unit.
That's why Google's sudden switch to Intel is troubling for ARM. Instead of upgrading from a Cortex-A9 design to a faster Cortex-A or ARMv8-A design, it went with Intel.
Since Google Glass isn't a commercial success, the change doesn't directly impact ARM's bottom line, but it damages its reputation. In the past, only a handful of second tier mobile players including Asus, Acer, Lenovo, and Dell installed Intel chips in their phones and tablets. But now, if other mobile device makers follow Google's example, ARM could start losing market share to Intel.
Why Google Glass matters to Google
Google Glass clearly isn't catching on as quickly as Google had hoped, due to social awkwardness and privacy issues.
Last May, Bite Interactive found that 90% of Americans didn't plan to buy Glass, which is currently available to "Explorers" for $1,500. Many developers, who had waited for nearly three years for Google to launch Glass commercially, have given up on the device. Meanwhile, action cameras and smartwatches have become more socially acceptable ways to replicate Glass' functions.
Google isn't giving up on its geeky gadget, however. It launched its "Glass at Work" initiative in June to focus on business uses for Glass. Glass has shown promise in the health care industry, but practical uses in other fields is still debatable. Nonetheless, Google profits from collecting information and tethering users to the cloud, so Glass remains a lucrative (albeit niche) way to do so.
Looking ahead, the success or failure of Google Glass won't make or break Intel, ARM, or Google. But the decision to use an Intel chip rather than an ARM-licensed one could be an early sign of coming changes in the industry.
Investors interested in the growth of the wearables market might want to consider investing in Intel instead of ARM. Intel isn't afraid to take temporary losses to steal market share away from ARM, but those losses could pave the way to bigger long-term gains in the future.