After AMD’s APU tech day in Beijing, AMD’s tablet focused SoC lacks any publicly committed hardware partners.

amd apu14 beijing 2 1024x576 Mullins can’t find design wins

Spencer Pan, AMD’s VP for China, opens APU14. (Photo: Chris Low/VR-Zone Chinese)

Blame it on Intel’s aggressive contra revenue stratagem, or AMD and its partners still not feeling ready to publicly release the chip, but after AMD’s Beijing tech day – which focused on the APU in the China market – its new mobile SoC still lacks any hardware to call its own.

An announcement of Mullins’ hardware wins was an expected highlight at AMD’s Beijing event. Mullins’ mobile cousin, Beema, had a number of hardware wins from a handful of vendors on display at the event’s product showcase but tablet wins were nowhere to be found (as a refresher: Mullins is geared towards tablets, whilst Beema is for notebooks and clamshell devices). Beema’s wins were present, but the form factors and system configurations will hardly resonate with savvy consumers.

“In terms of tablets, we haven’t been disclosing the design wins. But it’s fair to say that within [AMD] the focus is on the mainstream, value and premium clamshells,” AMD’s Bernd Lienhard told VR-Zone. “Where we are today with the products given the performance and graphics, that’s the best fit for us and we believe that’s where our value sits so that’s what we’re going to focus on.”

AMD didn’t elaborate more than that on the issue, but it’s not hard to see why OEMs are hesitant about placing AMD silicon in their tablets. Intel, with its contra revenue program, can simply undercut AMD by giving OEMs a better deal. It’s no secret; that’s the purpose of Intel’s $100 million development fund in Shenzhen. Whether this is a fair business practice is a debate that one day might happen in the courts, but until then that’s simply a market reality for AMD.

For AMD’s upcoming launch of mobile Kaveri, slated for Computex next month in Tapei, Lienhard took a similar stance on the possibility of hardware wins in the tablet space being announced. He confirmed there would be an announcement of hardware wins for mobile Kaveri in the embedded space, but when asked about the possibility of tablet wins being announced he simply answered “We can’t disclose this.”

Until then whether AMD is making any realistic progress on bringing its silicon in tablets into the market will be a known unknown for the public. AMD continues to play coy on whether serious announcements are on the way, but even if they come at Computex it’s not known how much market share they could possibly get as they will be Windows tablets — AMD won’t have anything Android compatible out for quite some time — which have historically had a lukewarm reception from consumers. Embedded wins are important for the company’s bottom line, but relying on these will force AMD into obscurity and these types of wins are rarely in the public’s eye or interest. The same goes for server wins: it’s a high-margin business but not consumer facing. AMD can do fine with the income, but it will have to settle for a lesser overall market shall and reduced corporate profile.

But for its part, AMD maintains the party line that it is engaging with OEMs.

“We have engaged with ‘the top three’ OEMs in the world to handpick some of the platforms in the tablet space, and we probably have to wait on them to announce it,” Lienhard said.

The big question is: why the wait? Beema and Mullins are simply process tweaked versions of Kabini and Temash with no architectural changes. They should not require a change in socket or motherboard designs. OEMs and ODMs have effectively had a year’s head start to design products that should be more or less compatible; AMD has had a year to convince hardware partners to align wins with the launch. Following this logic, it’s clear that the AMD is having a very hard time building confidence in its product.

But perhaps there will be an announcement next week that AMD can declare the new Microsoft Surface one of its hardware wins.

Photos by: Chris Low, VR-Zone Chinese