An Intel-MediaTek Deal Makes Sense, But is Not Likely

An outright takeover of MediaTek by Intel to give Intel a fighting chance in the mobile market is likely to occur in the next two-to-three years, if one analyst’s prediction is to be believed.

RBC Capital Markets Analyst Doug Freedman said in a research report released Wednesday that such a deal will happen as Intel struggles to organically increase its mobile market penetration organically while increasing shareholder value. For Intel, an acquisition of MediaTek — which doubled its baseband chip revenue between 2007 and 2013 from 7.3% to 14.4% — would give Intel a 30% market share within seven years compared to the 4% slice of the market it currently has that largely comes through a multi-billion dollar contra revenue program.

“Instead of Intel continuing to spend $4-$6 billion a year to enter the market (higher end of spending range as it achieves success), hypothetically, an acquisition of MediaTek may reallocate Intel’s best-in-class under-utilized fabs and financial resources to a rising star in the system-on-chip world, solidifying MediaTek’s market position,” Freedman wrote in a note to clients.

Freedman said that an acquisition of MediaTek, the idea itself has shades of Intel’s deal with Rockchip from earlier this year, would come in around the $27 billion range, which would be a premium of around 30% on the company’s current stock. Freedman says such a deal, while it comes with its own sticker shock, would create much more shareholder value in the long run than Intel trying to build the business up on its own.

The right ingredients, the wrong recipe

On paper, Freedman’s speculation of a possible deal makes some sense but is not likely in the long run because of practicalities. For its part Intel does need all the help it can get to increase its mobile market share, hence its deal with Rockchip, but an outright acquisition of MediaTek is still far fetched.

While MediaTek does have some very innovative IP in its portfolio, particularly in the space of heterogeneous computing (part of the reason why its a founding member of the HSA Foundation), its success is also the product of it being in the right place at the right time (read this author’s interview with MediaTek executives from earlier this year for more on this topic). Aside from HSA, MediaTek is best known for creating SoCs for low-cost devices marketed towards the developing world with CPU cores licensed from ARM and GPU cores mostly licensed from Imagination. MediaTek’s chips find their home in the devices being sold in tier-three and four cities in China, as well as growing regions south east Asia. MediaTek simply had the right experience and the right (with geographic proximity) supply chain allowing it to cash in on this market. Any firm could have had this success, if it were as ambitious as MediaTek.

This is where MediaTek’s value comes from. While Intel’s x86 platform is not quite at the competitive level as ARM for mobile applications, Intel would, under the right set of circumstances, be very capable of being “a MediaTek.”

So if Intel were to buy MediaTek it would be curious to see what it would do with it. There are too many unanswered questions to make the deal seem plausible at this time: would it be MediaTek powered by Intel, with Intel’s CPU and GPU IP replacing ARM’s and Imagination’s? How would customers react? What about chip pricing?

In addition if Intel purchased MediaTek it would find itself in an awkward situation, as it would now be cooperating heavily with rival AMD in developing the HSA standard.

Freedman has been shopping this theory around for some time, with it first appearing last November after Intel’s analyst day — an event that left many with a sour taste in their mouths. The theory itself is interesting, and worth considering, but don’t count on it happening anytime soon.