With Intel’s tablet business taking off, why is its mobile unit revenue plummeting?
The company saw a significant spike in tablets using its chips — 10 million last quarter — but its mobile revenue was just $51 million — down 83 percent from a year earlier and by more than two-thirds from the prior quarter.
Yes, its phone modem chip business is tailing off, but that alone doesn’t explain such a steep decline. The reason turns out to be something known as “contra revenue.”
Despite the name, it has nothing to do with funding Nicaraguan rebels. Rather, it is financial jargon that means Intel is essentially paying tablet device makers to use its chips.
For 2014 anyway, Intel is selling a chip into low-end tablets that was never intended for that market. As a result, the chip is more costly and complex to design into devices than rivals. Because tablet makers would otherwise choose the chip leading to an overall less expensive design, Intel is paying the manufacturers who use its chips to offset the higher costs that an Intel design entails. It is also paying some of the device maker’s costs for designing the Intel-based tablets.
Although not unheard of in the tech world, it’s an unusual business model that only a desperate and wealthy company can afford.
Intel has been upfront that this was its plan — it disclosed this at a November 2013 analyst meeting — but it showed up quite prominently in the latest earnings report as Intel’s tablet volume started to take off.
And, of course, since Intel says it is “on track” to reach its goal of selling 40 million tablet processors this year, that means more huge losses for the quarters ahead. Indeed, the more “successful” Intel is at getting device makers to use its chips, the more money it will lose.
The company has said that the tablet program is expected to take the company’s entire profit margin down by as much as 1.5 percentage points this year. Considering Intel has tens of billions of dollars in annual revenue, that means Intel is racking up hundreds of millions of dollars in expenses.
Intel finds itself in this position because it originally envisioned the Bay Trail chip it is selling to tablet makers would wind up in high-end devices, iPad-like rivals that sell for several hundred dollars apiece. Instead, Intel’s opportunities have been in lower-end devices such as the Asus Memo Pad, a device that costs around $150.
The bet is that next year Intel will be able to keep some of these tablet customers when it has more integrated chips that can fit in lower-end tablets without requiring such outlays from Intel.
Intel doesn’t expect the mobile unit to turn profitable next year, but the losses should narrow, CFO Stacy Smith said in a conference call with analysts on Tuesday.
“It won’t be profitable, but we should be able to improve it nicely,” Smith said.
CEO Brian Krzanich put a finer point on things, assuring analysts that the investments were both temporary and necessary.
“I mean, clearly, we don’t go into businesses to lose money, and we believe that over time we can make this a profitable business,” he said.
To do so sooner, Intel is making another big bet. The company, for the first time in its history, is making Intel-architecture chips outside its own factories. The chips, known as SoFIA, take a design inherited from the Infineon acquisition, but swap in an Intel processor core for the ARM chip that Infineon had used.
Over time, Intel plans to move the chip’s production in house, but sticking with an outside manufacturer for now will get it to market faster.
In the meantime, Intel’s losses are clearly mounting.
This article originally appeared on Recode.net.