China is the critical territory for all mobile device and chip vendors, not just for its own vast size and growth potential, but because it is the guiding force for other emerging mobile economies. India and Brazil are creating their own device ecosystems too, but these are heavily influenced by China, whose companies are driving aggressively into the new powerhouses. Bold strategies are required if established handset and system-on-chip (SoC) vendors are not to lose their grip. Nokia and Motorola, once dominant and China and India, were usurped by Samsung and Apple largely because of their failure to crack the redefined US smartphone market. Now their near-duopoly is being eroded and they may suffer the same pattern in reverse, with China the deciding factor this time, and rivals like Lenovo and Xiaomi set to profit.
The same goes for the chip companies. Leading in China and India requires a new set of economics, and like Apple in smartphones, western chip giants like Qualcomm have to choose between their traditional generous margins and maintaining market share. Qualcomm knows that it is no longer competing primarily against Broadcom and Nvidia but against MediaTek, Spreadtrum and other rising stars from greater China. Being the adaptable company it is, it has already taken steps to address the challenge with its Qualcomm Reference Design program. This provides low cost, out-of-the-box platforms for manufacturers to create handsets and cheap tablets easily and affordably. But just as importantly, it brings low cost Chinese players into Qualcomm’s usually rarefied orbit.
But just a week after it was showing off the latest fruits of these alliances in China, Qualcomm finds that Intel, not the most obvious threat in ultra-low cost arenas, has made its boldest move yet to push its Atom SoCs into the mobile mass market.
Intel has thrown down its biggest challenge yet to ARM’s mobile ecosystem, announcing an alliance with China’s Rockchip to push x86-based chips into the mass market for the first time, and making a bold break with its past models.
There has always been a strong logic to Intel giving up the low end system-on-chip (SoC) as a lost cause and concentrating on premium devices which would require the performance of its architecture, and which would eventually replace its PC base ‘ larger-screened tablets, ultrabooks, Chromebooks and so on. But the firm has clung to the goal of achieving mobile mass, and an important foothold in emerging markets, despite numerous setbacks. Indeed, the first smartphones to be powered by its silicon were low end designs from Indian vendors. But it has been held back by its lack of fully integrated SoCs and by an ecosystem which was not geared to commodity consumer gadgets.
The key market in which to try to change all that, of course, is China. The company has had plenty of Chinese alliances and initiatives in the past, but not one focused so specifically on low end mobile devices, in this case cheap Android tablets. It has three points of significance beyond the immediate shipments potential ‘ if successful, it will be a powerful proof point that the x86 architecture really can compete with ARM in terms of cost and power consumption, after many years of trying; it will open up new ecosystem partners to distributed x86 designs (vital in China’s fragmented and often inward-looking market); and most importantly of all, it indicates CEO Brian Krzanich’s willingness to embrace new business models in pursuit of success in mobile and internet of things (IoT) segments ‘ identified, when he took the helm last year, as critical for future growth.
He did not just promise to redouble Intel’s efforts to become a mobile big hitter at last, but threw some impressive new thinking behind the goal. He is turning his firm into a foundry supplier, applying its redoubtable expertise in semiconductor processes to a wider set of revenue streams. And with the Quark IoT processors, and now with the Rockchip deal, Intel is opening up its architecture for third parties to adapt, in order to target a broader range of territories and product categories, some of them tough for the giant to pursue directly. It may be just a step away from taking on ARM at its own game and licensing the x86 and Intel Architecture IP, or elements of it, completely.
That is not a feature of this latest deal, but it is important in its own right and likely to be the blueprint for others. Rockchip is a fabless semiconductor manufacturer based in China and currently develops ARM-based mobile SoCs, mainly for low end devices. But now it will work with Intel to create an x86 product, and just as significantly, help sell it into its network of local customers and contacts, targeting vendors and product specs which would normally be off Intel’s radar. ‘By partnering with them, we’re getting a product out that’s targeted very specifically for these tablet designs that wouldn’t probably otherwise get onto our roadmap,’ Krzanich said in a statement. ‘They’re bringing a speed and execution and proliferation that would otherwise not get done.’
In the fourth quarter of 2013, Rockchip sold 9m tablet processors, of a total of 88.3m shipped ‘ about 50% more than Intel did, according to IDC calculations. It sold 40m mobile app processors in total last year, about the same as Allwinner, though it was weaker in tablet chips. Intel is targeting shipments of 40m tablet SoCs this year, but only did 5m in Q1, so extending its reach in volume markets will be crucial.
By the middle of next year, the new partners say they will launch an entry level, Intel-branded SoC for low end Android tablets. This will combine four Atom cores and an Intel 3G modem, and will fit into the US firm’s SoFIA family, unveiled at Mobile World Congress.
SoFIA will use the Silvermont core and will be made in 22nm technology, and will be Intel’s first offering with a fully integrated modem ‘ the firm admits the lack of that capability, something in which Qualcomm excels, has been its biggest barrier to low end device success. With this Rockchip variant, SoFIA will have three products by the third quarter of next year ‘ the others are a dual-core/3G part, launching in the second half of this year, and a quad-core/LTE version, due in the first half of next year.
The combination of four cores with the older modem highlights some important differences between the Chinese market and others ‘ differences that often end up favoring the local suppliers, which often work more closely with the operators and understand their requirements. Chinese consumers are certainly demanding high powered smartphones, even at low prices, and that demand saw Qualcomm launch 64-bit and some other performance advances in SoCs for China, before the rest of the world. However, 4G is nascent and fragmented, despite China Mobile’s huge TD-LTE push, so in 2014-15, much of the growth in commodity tablets and other devices will come from the still-growing 3G networks. It is not clear which 3G variant(s) will be supported ‘ China Unicom’s W-CDMA seems certain, but Intel could also branch out into modems for China Telecom’s CDMA2000 or China Mobile’s TD-SCDMA.
The interesting aspect of the deal is the nature of the cooperation. Intel will provide the basic SoC and Rockchip will then integrate and design a special variant, optimized for tablets for its own markets. This is a departure for both companies, but shows Intel’s greater willingness to adopt new models in order to expand its markets. The deal is “designed specifically to expand the breadth of and accelerate the rate at which we bring our Intel architecture and communications solutions to more customers and segments of the global market for entry and value tablets,” said Kari Aakre, an Intel spokeswoman.
It will gain a better cost base and improved local ecosystem contacts through Rockchip, and also, by making its platform more flexible, will start to take on ARM at its own game ‘ with an architecture that, even though not yet fully licensed, can be adapted to a customer’s specific requirements. The chip will initially be manufactured by TSMC, rather than Intel itself, and like the other SoFIA products, transitioned to Intel’s own fabs, probably in 2016.
There are no signs the deal is exclusive, so Rockchip will presumably continue with its existing ARM-based mobile ranges, but it will gain far greater differentiation over its various competitors by harnessing the rarely shared Intel technology and being able to add its own twist to it. Its arch-rival Allwinner does not currently have access to integrated 3G/processor technology, for instance.
Now that Atom has overcome many of its old problems with power consumption, it can be a viable platform for fully mobile devices, but it lacks the ecosystem of ARM ‘ of course Intel is a huge power base in its own right, but not in mobile, especially outside the post-PC categories such as large tablets.
This could be the first step on the way to building an ARM-like family of partners to take its platform into all kinds of new markets which Intel finds it hard to penetrate directly, for all kinds of reasons of cost and history. Intel has hinted at a similar change of direction for the internet of things, which similarly has completely different structures and cost bases from those to which the giant is accustomed. Quark has an ‘open architecture’, with Intel offering ‘fully synthesizable’ chips. These have hooks in the silicon to add others’ IP blocks, to customize designs for specific tasks. But a bolder move, to license all the IP, could follow, in a world where IP and common platforms are the route to power and profit, not churning out every low margin chip and sensor.
‘We are always looking for innovative ways to differentiate our product portfolio, and the first-of-its-kind collaboration with Intel helps us do this,’ said Min Li, Rockchip’s CEO, in a statement. ‘The combination of Intel’s leading architecture and modem technology with our leading mobile design capability brings greater choice to the growing global market for mobile devices in the entry and value segments.’
The new handset landscape
Intel’s new partnership is not, initially at least, about handsets, but the firm can be expected to build on the template and push SoFIA into smartphones and into other relationships in future. It will certainly need to cultivate key Chinese device makers, as well as ecosystem players like Rockchip, if it is serious about growth at the low end.
And it is not just in China and emerging economies where the big two are feeling the pressure. The latest figures from research firm Kantar Worldpanel ComTech indicate that the traditionalist European smartphone market is fragmenting, with Apple and Samsung being squeezed not only by a recovery in some older brands like Sony and Motorola (soon to be Chinese itself of course), but the rise of Huawei and Lenovo. There is also another interesting trend ‘ the revival of strong local European phonemakers, particularly Wiko Mobile in France, which is now the country’s third largest smartphone vendor. This is a pattern not seen for the best part of a decade and shows how SoC makers will have a longer list of companies to target. That will be seen as positive, because Apple and Samsung increasingly make key components themselves or rely on a small set of vendors, but big silicon players may need to adapt their sales structures and pricing to beat low cost Chinese rivals to win slots with the new breed of smartphone challengers.
Dominic Sunnebo, strategic insight director at Kantar, explained some of the behavior which is starting to change the smartphone landscape. “Consumers are starting to realize the true cost of handsets, and as a result they are shopping around to find cheaper alternatives,’ he said, referring to the falling reliance on subsidies in many areas.’This shift in behavior plays directly into the hands of lesser known brands like Huawei and Wiko, who are able to offer competitive technologies for a fraction of the price.’ And with Apple, as yet, failing to come up with a really convincing strategy for the low cost market, Android can only improve its position as the driver of those affordable products.
Nvidia’s CEO is also taking new routes, away from handsets
While Intel is determined to play in this fragmented, low end game ‘ in some cases by joining forces with the budget chipmakers rather than trying to out do them ‘ other companies will remain focused on the profits of the premium segments. Nvidia admitted recently that it had effectively failed in the handset space and was looking for alternative sources of growth for its GPUs (graphics processing units), especially in high end categories.
Like Intel, Nvidia is suffering from the falling levels of PC sales and needs to compensate for that with strong presence in post-PC devices or in entirely new markets. With its Tegra mobile processor and graphics platform, it initially made a big splash in tablets by moving early, ahead of Qualcomm, in high performance areas like quad-core and 64-bit. It powered the first two models of the Microsoft Surface RT and other devices, though its lack of an integrated processor/modem solution, and Qualcomm’s dominance, held it back in handsets. It acquired innovative 4G modem designer Icera to put this right, and has just started shipping all-in-one offerings ‘ just as it seems to be defocusing on smartphones, which may raise question marks over the Icera technology.
In an interview with CNET, CEO Jen-Hsun Huang said Nvidia would now only focus on smartphones and tablets in which high end graphics were the key differentiator ‘ a small subset, he acknowledged. He admitted that the integrated product, Tegra 4i, ‘wasn’t that successful for us’, despite its technical prowess. He explained: ‘I would say that when we first started this, we thought that bringing 4G to entry level phones, mainstream phones, integrated with our apps processor would be a real opportunity. I think that the phone marketplace has commoditized really, really fast. It is not our strategy to go after commodity phones.’
Like some device makers which are looking like victims of the handset bloodbath, such as BlackBerry, Nvidia sees its salvation in emerging connected device categories such as cars. Huang said: ‘Mobile technology is revolutionizing [products] all over the place. It’s going to revolutionize TVs. It’s going to revolutionize, obviously, computers. But it’s also going to revolutionize games and cars and all kinds of stuff. Our strategy is to go and focus our energy on segments of mobile where we can add the most value. The area where we can add the most value is visual computing, graphics, GPUs, etc.’
That sees Nvidia very much retreating to its area of strength, and putting its focus on gaming ‘ especially as played on large-screened devices like TVs and car displays, not handsets. ‘We’re going to focus on car computers where visual computing and supercomputing technology is really important to the future of cars. And then we’re going to focus our energy on segments of mobile computing, segments of phones and tablets where computer graphics are really important,’ said Huang.
The failures at Nvidia have been all about timing, not product capability. It arguably put too much resource into stealing a march on Qualcomm in multicore, before many OEMs really wanted that. Then, as performance demands did mount, it delayed Tegra in order to get Tegra 4i finished, only to find that the smartphone space had commoditized far more quickly than it had anticipated.
Huang said: ‘I would say that Tegra 4i didn’t pan out. We learned a lot in the process. Look, we built a great chip. LG’s shipping it in the rest of the world outside the United States. It’s a fantastic processor. But from a business strategy, it wasn’t a success. So I learned a lot from it. It’s OK. I’m glad I did it, and now we’re moving on … but I think that I learned we are not a commodity player. Going after something that starts with mainstream should have discouraged us. At the time, it looked like our 4G solution was quite far ahead of the market, but that window’s closing very quickly. We’re just not a cost player.’
With Microsoft reportedly turning to Qualcomm for its next ARM-powered Surface tablet, Nvidia is likely making a wise decision to turn to more rarefied markets, but there is a question mark over how far it will adapt its modem technology for cars and gaming platforms ‘ and therefore where the future of the former Icera lies ‘ and also whether there is sufficient volume in its new chosen targets. Like others, including Intel, it is relying on creating a cloud platform behind its device ecosystem, and here it does have some interesting credentials because of its expertise in supercomputing processors. It will never be free of Intel’s shadow in any of its strategies, but at least it can stop worrying about MediaTek now.