Two and a half years ago, Intel dropped a bombshell. The company’s products were struggling to find their footing in the mobile market and Intel’s smartphone and tablet division wasn’t performing as desired. To address this, Intel would partner with rival TSMC to build smartphone and tablet chips on the latter’s 28nm process technology. Initially, these devices would be limited to 3G modems and budget products, but Intel planned to migrate to 14nm hardware and 4G modems in fairly short order.
According to an article by Patrick Moorhead of Moor Insights & Strategy, that entire plan is now dead. The entire SoFIA project is canceled, in all flavors, as is Intel’s plan to launch its Broxton SoC for smartphones and tablets. Broxton would’ve been based on the Goldmont CPU core — Intel’s first major update to Atom since 2013. We reached out to Intel, which confirmed the news: SoFIA 3GX, SoFIA LTE, SoFIA LTE2, and Broxton have all been canceled.
On the one hand, Intel’s announcements today aren’t a dramatic surprise. Here’s the roadmap Intel initially laid out for SoFIA and its follow-up family of products:
Note that Broxton was originally supposed to be released a year ago. SoFIA 3G was supposed to be ready for the back half of 2014 and SoFIA LTE would launch in 2015. In 2015, the updated slide looked like this:
Instead, it looks as though neither the SoFIA 3G or the SoFIA LTE processors ever actually came to market (they’re both listed as “Announced” as opposed to “Launched” at Ark.Intel.com). Only the SoFIA 3G R processors are listed as actually launched (these may be related to the Intel / Rockchip team-up we discussed last year).
The death of both Broxton and SoFIA implies that Intel is getting out of the tablet business altogether. Pat Moorhead believes they’ve got a better chance at staking a claim to 5G than they did playing catch-up with 4G, and that’s certainly possible, though they’ll face entrenched competition from companies like Qualcomm, whose near-lock on 4G modems helped propel the company to its current market position.
The memo CEO Brian Krzanich published earlier this week might have demoted the PC business to just another thing connected to the Internet, but he at least mentioned it. He had nothing to say about Intel’s tablet and smartphone businesses at all.
One interesting aspect to these changes and pullbacks is how dramatically conventional wisdom has shifted over just the past five years. When Microsoft announced that Windows 8 would be compatible with ARM processors and Intel declared it would build mobile devices based on Atom, these announcements were hailed as absolutely critical to the future of both companies. Microsoft and Intel, the thinking went, absolutely needed to be in mobile (and ready to run the desktops of the future on non-x86 hardware). AMD, meanwhile, was often chided for failing to launch itself into these markets or compete in them, despite having drastically fewer resources to its name.
Fast forward to the modern day, and Windows Phone is guttering and on the verge of total failure while Intel is in wholesale retreat from the phone and tablet market. After billions of dollars of investment, all the two companies have to show for their efforts is a fraction of the market and an unprofitable contra-revenue strategy that never turned into a sustained advantage. From Android-x86 to Windows RT, the bad bets piled up on each other, year after year.
Not that AMD is in great shape these days — but things would be even worse if it had tried to shove its way into mobile.