Significance: Texas Instruments has made two significant changes of direction this year. It has extended its KeyStone ARM/DSP system-on-chip architecture well beyond its base station roots and into all aspects of cloud infrastructure, most recently high performance servers. Here it will look to establish its platform ‘ and by extension the ARM processor architecture ‘ as the standard for intensive floating point tasks, even while leaving x86 intact as the central processor. It is pushing new approaches to the server chip, including integrated networking and DSPs, in many cases ahead of Intel. The significant expansion of KeyStone’s target applications indicates that future growth will often lie outside the mobile network, despite the continuing rise in wireless build-outs and traffic. The requirements of a server, a router, a base station and an industrial piece of equipment are far more similar than they used to be, leading TI, Intel and Freescale to talk up platforms which can span them all. The focus is on data processing, not specifically on mobility any longer, especially as the base station sector becomes commoditized by the move to small cells. Parallel trends are seen on the device side of the mobile equation, where chip vendors are looking beyond the handset for new opportunities. Nvidia’s recent results suggest that the tablet is the most fertile ground, but that is likely to prove a red herring. By existing the tablet and smartphone processor space and refocusing OMAP on the embedded market, TI may look hasty, but in fact it has recognized that mobile apps processors will be hit by rising competition, price wars and consolidation. The returns in the non-cellco world may be slower, but the embedded and cloud markets are increasingly where companies will innovate and find a differentiated play.
Nvidia should watch TI’s tablet exit with trepidation as well as hope
Tablets have been portrayed as the single-handed savior of the device chip providers, compensating for the slow but sure decline of the conventional PC, and adding a higher margin product for mobile suppliers facing smartphone commoditization. However, this is a heavy burden to place on a form factor which is still in its infancy, despite strong early growth, and which delivered volumes of about 28m units in the third quarter, less than a sixth of the number of smartphones (181.1m according to IDC, and still growing healthily), and less than a third of total PCs. Is there really sufficient volume in tablets to support the growth ambitions of the smartphone players and the PC chip suppliers combined?
Some mobile giants certainly think not. Texas Instruments is exiting the tablet and smartphone processor markets altogether and seeking growth in analog products and a major expansion in its infrastructure play (see separate item). This is a move which will bring short term hope to Nvidia, which is the poster child for a tablet-based growth burst, and which could target TI’s flagship slate, the Kindle Fire. But Nvidia should eye TI’s move with fear too, as a warning not to place too many bets on the tablet.
Like TI, Freescale is more focused on pushing its processors into embedded sectors, while STMicro has recently been more bullish about this business too, especially when compared to its mobile joint venture, ST-Ericsson. Nvidia is currently riding high on tablet app processors, but to put this at the center of its strategy is risky, because it faces being squeezed on three fronts ‘ by Qualcomm with its integrated Snapdragon platform and mobile might; by companies moving down from the PC space, notably Intel; and by those targeting low cost tablets and ereaders, which will include a host of Chinese and Taiwanese chip and module makers in future, as well as Freescale.
Of course, much depends on how a tablet is classified. PC decline may be slowed or reversed if new form factors such as ultrabooks take off, but though these have clear roots in the PC ecosystem, they will increasingly sport touchscreens and ‘convertible’ modes between slate and keyboard usage. Some of the devices proposed for Windows 8, in both its flavors, are hard to categorize as to whether they are tablets or new-style PCs. Increasingly, the ‘pure’ tablet will be a low cost item ‘ below 8-inches, monolithic, highly portable. The blueprint is the Kindle Fire and the vendors will either target low cost markets and high volumes (mainly an ODM play) or, like Amazon, use a cheap tablet to drive additional content sales. In both cases, the margins for the processor suppliers will be thin. The higher returns will lie in the PC/tablet center ground, in form factors which combine elements of both devices and target the higher value mobile users currently chased by larger slates and ultrabooks. These products will take on the roles of the notebook, the TV second screen and the web browser and will represent a significant market opportunity in size and price. However, there will be many bets taken on failed form factors before the sector stabilizes, and when the successful vendors emerge, there will be huge competition for their slots ‘ chip designers from the PC and smartphone arenas combined, plus emerging new names from China.
No wonder TI is backing away, though its timing may be premature. Timing has not been the firm’s strong point, at least on the device side, in recent years. It took a bet on smartphone makers wanting standalone apps processors a few years ago when it pulled out of merchant basebands, only to see rising demand for platforms like Qualcomm’s, which tightly integrate the modem and processor ‘ and CPU-only vendors like Intel and Nvidia rushing to acquire baseband assets (Infineon Wireless and Icera respectively).
The loss of an integrated offering is one factor which has weakened TI competitively in the mobile processor sector it helped to define with its OMAP platform. When it first indicated it would back away from smartphone processors, back in September, it was understood to be seeking a buyer for OMAP, with Amazon rumored to be one candidate, in a bid to control its own platform, Apple-style. However, it seems that no buyer was found, and TI says it will refocus OMAP on the embedded market, a shift already seen ‘ though with a less established mobile CPU ‘ at Freescale.
This week, TI confirmed the strategy and announced it would cost 1,700 jobs. Both the decision and the scale of the job cuts was unusual for the firm, which tends to move more gradually, and some analysts believe it should have stuck around in the tablet/smartphone space for a bit longer, given that OMAP is well established and there is a window of opportunity in tablets before Intel gets serious ‘ a window being exploited effectively by Nvidia, whose tablet chip business doubled in its third quarter.
Nvidia outdid Wall Street forecasts in Q3, crediting strong performance of its Tegra 3 processor in the tablet market. It reported net income of $209.1m, or 33 cents a share, up from $178.3m, or 29 cents, a year earlier, while revenue climbed 13% year-on-year to $1.2bn. It told analysts it was offsetting the decline in the PC market ‘ where it specializes in graphics processors ‘ with growth in apps/graphics processors for Android tablets, and forthcoming launches of slates running the new Windows RT operating system for ARM-based platforms. In more traditional sectors, Nvidia is also stealing market share from AMD in PC graphics, said analysts.
Such performance suggests that the tablet space, while still fairly small, is not yet as fiercely competitive as the handset market, and could have delivered returns for OMAP for a while longer. Nvidia will have keen eyes on the Kindle Fire, which uses OMAP. However, in the medium term TI’s decision seems a wise one. Nvidia’s strong figures disguise a fragile foundation for the tablet success, and high risks if the firm does not broaden its customer base (which may be tough to do as competition mounts and Intel, on the back of Windows 8, renews its slate assault). The bulk of its Tegra sales in the quarter were for two models, the Google Nexus 7 and Microsoft Surface, and early indications are that the latter will not have the mass market impact of a Galaxy Tab or Kindle Fire. Nvidia has largely failed to hit Qualcomm in smartphones, though it says it plans to launch a new chip next year optimized for handsets. Its acquisition of baseband maker Icera will enable it to launch an integrated processor like Qualcomm’s, though this will be in a crowded segment.
Another risk factor is the increasing dominance of Apple and Samsung in mobile devices ‘ they control over half the smartphone space, but both rely on their own processors. This was cited by TI as one reason behind its change of strategy, with executives pointing out that ‘large customers are increasingly developing their own custom chips’.
Meanwhile, TI will retain positions in the smartphone and tablet in the shape of analog chips ‘ its primary focus these days on the device side ‘ and connectivity combo parts supporting Wi-Fi, Bluetooth and other links. As far as processors are concerned, though, embedded markets are the new promised land. “TI…will reduce costs and focus investments in its wireless business on embedded markets with greater potential for sustainable growth and …. long lifecycles,’ said the official statement.
In this, TI is following a far broader industry movement, in terms of growth, value and power plays, towards the internet of things. In a recent interview with Reuters, Greg Delagi, SVP for embedded processing, said growth in the embedded area would be slower than in mobile, but would “generate a more stable, profitable long term business”, targeting gross margins of 55% to 60% (in an uncertain timeframe). As with its huge bet on the analog chip space, this shows TI putting all its resources into areas where it has natural leadership, rather than forcing itself to be a challenger. It is already the leading semiconductor maker in the industrial chip sector, according to IHS iSuppli. Importing its mobile experience to some of those bases makes a lot of sense, as the embedded systems in cars and homes become more sophisticated, following handsets along the road to incorporating computing power and browsers. The industrial market is a fragmented one ‘ TI may have the lead, but its share was only 7.3% last year (up from 6.4% in 2010). Offering more options, and higher value systems, to key customers could help increase that further, and there is plenty of growth to be had ‘ TI’s revenue from the industrial business rose by 25% to $2.2bn in 2011, Delagi said.
Here, TI will meet different competitors, such as Infineon and NXP, though mobile specialists like Qualcomm are also ramping up their IoT challenge. Another company with high hopes for a new kind of mobile device is STMicro, whose smartphone processor and baseband platform is struggling within the ailing ST-Ericsson joint venture. But the Swiss parent continues to develop new products to give it a role in emerging mobile device categories, and the myriad chips which will power the internet of things.
For instance, STMicro has partnered with Xsens, a specialist in 3D motion tracking technology, to demonstrate the world’s first wearable wireless 3D body motion tracking system based on consumer grade MEMS sensors. The system combines Xsens’ sensor fusion algorithms and wireless protocols with ST’s iNEMO-M1, which incorporates MEMS motion sensors and the
STM32 microprocessor.
After working on similar technologies for ultra-real gaming and for movies, STMicro is turning its attention to broader consumer applications. Benedetto Vigna, general manager of the firm’s analog MEMS and sensors group, said: “Xsens’ unique 3D body motion tracking technology will help our customers innovate and deliver new solutions to their customers in sports, fitness, healthcare and gaming. We are at the dawn of a new era for Motion MEMS and we want to extend our leadership.”
Applications like this, as well as advances in low power consumption, will increasingly be the focus of real innovation and differentiation in the mobile world, rather than conventional processors in which squeezing in more cores and integrated graphics will be the main trends. TI has decided to opt out of that particular race, and it will not be the last. Tablets in the pure sense will have a ceiling in terms of volume, and an even lower one in price, while the broader hybrid device space will have high growth potential, but will consolidate rapidly among a few giants. Out in the internet of things, there will be far greater opportunities to come up with something unique and to have more than a small window in which to grasp a lead.
Texas Instruments pushes Keystone towards supercomputers
When Texas Instruments unveiled its Keystone system-on-chip architecture, it was clear that the brand new platform was designed to expand TI’s reach beyond its traditional stronghold in base stations. How far it aims to stretch is more surprising though ‘ after predictably pushing further into network processors and small cells earlier this year, it is now targeting servers as well as a wide range of networking gear with its combination of multicore ARM processors, DSPs and accelerators.
Like Freescale, TI is looking to provide a single architecture and software base which run the whole RAN gamut from metrocell to macrocell, and then beyond the base station to the network server, which has traditionally been a separate segment, and now to the general purpose server. When it unveiled the second generation of its KeyStone architecture earlier this year, it boasted up to four ARM Cortex A15 cores and multiple C66x DSPs, but perhaps more significantly, hardware acceleration for switching, packet processing and media. This inclusion of packet processing and switching right on the chip makes KeyStone more suited to data-driven computing platforms, and will help further two aims close to TI’s heart ‘ bringing the DSP into the server, and integrating networking on-chip. In the latter case, integrating up to five 10-Gigabit Ethernet ports on the chip not only supports intensive real time information processing and movement, but could do away with the need for a top-of-rack switch.
Already this year, KeyStone SoCs have been unveiled targeting the Cloud-RAN, network server and embedded equipment. The latest members of the family add to the options in the networking and embedded spaces, and add models targeted squarely at servers, as TI seeks to extend its ARM/DSP/network architecture to every kind of equipment that is used in the cloud.
Marvell and start-up Calxeda have led the way into using ARM-based processors in servers, mainly addressing low power blade servers for cloud providers. TI has stated it is not primarily targeting the direct replacement of server blades, but instead is looking at the trend for large data centers to turn away from monolithic server designs in favor of ‘breaking jobs down and putting some things on a DSP or A15’, as Tom Flanagan, director of technical strategy at TI’s wireless unit, puts it. Power efficiency is also critical to these large centers, and TI says it can deliver as much as 10 times the MIPS per power as x86 in a heavy number crunching environment. Therefore, while the shift to cloud architectures is TI’s key driver, it is looking right up to high performance servers and supercomputers, where ‘ rather than seeking to oust the incumbent Intel x86 architecture – it will offer chips to offload heavy duty floating point tasks. In this role, it will take on Nvidia, which makes a similar play, though using its graphics processing units rather than a DSP-based solution like TI’s. Target functions would include media processing, video analytics and industrial imaging and control.
The latest Keystone chips, most of which are going into production now, consist of six 28nm SoCs based on ARM Cortex A15 CPUs, C66x DSPs plus security and networking blocks. As well as servers, they will target routers and switches, opening another front in TI’s war against Freescale, as well as challenging Cavium and LSI. The last of these has been the flagwaver for using ARM cores in the network infrastructure market, while Freescale also recently gave ARM a significant boost by including the Cortex platform in its PowerPC-based base station range ‘ though, for now, only at the low end.
Flanagan believes the open ecosystem of ARM will ensure its success in many markets, especially if chip vendors stick with a unified approach ‘ unlike some ARM users such as Qualcomm and Apple, TI uses the standard core design rather than adapting it, and relies on its accelerator and software efficiency to deliver a performance differential. The biggest of the new chips use up to four A15 cores and six DSPs to deliver up to 352GMACs and 19,600 Dhrystone MIPS.
Distinguishing features of its Keystone chips include a 256-bit interface from the cores to the SoC, clocked at the full 1.4GHz data rate of the cores, while some rivals use 128-bit interfaces clocked at only 33% to 50% of that rate. TI also squeezes up to 18Mbytes of aggregate memory on its top end SoCs and says it is unique in supporting Gigabit and 10Gbit Ethernet MACs.
The last of these features is critical to its positioning. Integrating networking onto the chip is a popular topic among chip designers as servers need to take on far greater levels of real time data analysis and other tasks where information must be processed in parallel and at ultra-high speeds. However, TI is rolling out 10Gbit capabilities imminently while Intel, for instance, is talking about 100Gbit networks-on-a-chip, but with no timelines. But adding integrated networking to a DSP/ARM architecture, in real world products, may allow TI to succeed in its aim of getting DSPs into servers.
While the high end server models will get the most attention, only two are devoted to that space. Another two come with just one DSP core and are aimed at routers and other wireless infrastructure plus cloud platforms. And two low end variants will focus on embedded and industrial products such as traffic or surveillance cameras, or VoIP, and come with no DSPs. These build on an earlier decision by TI, discussed back in March, to push KeyStone into the embedded world. Then, it release three chips for devices in sectors like industrial automation, imaging, video surveillance, medical, and audiovisual infrastructure. These are the dual-core TMS320C6657, which runs on two 1.25GHz DSP cores, delivering up to 40GFLOPs, and the C6655 and C6654, single-core parts providing up to 20GLOPS and 13.6GLOPS, respectively.
Among the new offerings, the lower end chips will ship in the second half of next year while their high end stablemates are sampling from next month. While lacking the glamor of the supercomputer chips, the networking infrastructure and the industrial sensor-based chips are aimed at markets where TI has strong roots, and so is trying to win more slots with well understood customers, rather than attacking a new space. But it is in servers where the biggest spoils are available, and where TI can finally make the concept of DSPs inside servers, which it has talked up for two or three years, a commercial reality ‘ not to mention supporting another key concept, fully integrated networking on a chip, well ahead of its noisiest exponent, Intel. As in mobile device processors, where TI is backing away from the tablet rat race just as so many players are entering it, in servers it is avoiding the most fashionable approach, the Calxeda-style cloud platform, and eyeing a somewhat different twist on the modern server. As such, for all its shift towards analog businesses, it may find a new lease of life for its DSP activities too.