Your browser is not supported. Please update it.

2 July 2008

Analysis: Chip majors in danger of missing Chinese mobile TV market

Major global chipmakers are in real danger of missing out on yet another huge Chinese market, for its homegrown mobile TV technology, CMMB STiMi.

The same thing happened in Japan, when doubts about the viability of the ISDB-T technology meant that none of the major chipmakers built a base in supplying ISDB-T chips, and now have missed out on a 25m portable device market in that country, plus a further similar number likely in Brazil, and the potential that other parts of South America will also follow the Brazilians. Instead, local companies like Sharp, Megachip and Fujitsu have sewn up this market.

So far, tiny Israeli chipmaker Siano Mobile Silicon reckons it is the only outsider that is making chips for the CMMB STiMi local system which will be launched, China hopes, just prior to the Beijing Olympics this August. Alon Ironi, CEO of Siano, told our sister service Faultline last week: ‘We have had silicon for about a month for the final version of the chip, but we didn’t want to announce it until we’d had some field trials. The reception is good in both Beijing and Shanghai, and the system works and even gives good indoor reception.’

The Siano chip masks some difficult complexities. The Chinese system has been designed with two signals, a 2.6GHz satellite signal and a 700MHz terrestrial signal, and regulator SARFT will not license a device unless it can pick up both signals. This is tricky because at present there is no signal coming from any satellite and there won’t be until the early part of 2009, that’s assuming that the satellite goes up on schedule. EchoStar, the primary provider of S-band satellite capacity for the project, quietly revealed in an SEC in late May that it was suspending construction of the CMBStar satellite. (China Satellite Mobile Broadcast, a company overseen by SARFT, last year selected CMBSat, a Hong Kong-based affiliate of EchoStar, as its partner.)

However, recent discussions about the delayed satellite have been held between EchoStar, satellite builder Loral and SARFT, which suggests that the agreement over the provision of the satellite is back in negotiation, even though it is half built. There are still likely to be delays, and potentially SARFT may decide to commission another satellite, but meanwhile all devices need to be capable of receiving both signals, and although cross-spectrum diversity is not part of the specification, it is one way to make the signal clearer on devices.

‘We have two chips, one that will support two of each of the antennas, and one that will only support one of each. When we use two antennas with the satellite signal, we found that there was a gain of 3dB to 4dB, which is the difference between strong in-building reception and poor reception,’ said Ironi.

But the Siano chip is architected so that it samples the quality of the signal on all four antennas, and picks the strongest two signals and then combines them logically on the chip. This means that diversity benefits can be achieved using one satellite signal and one terrestrial, or both satellite or both terrestrial.

‘We are not expecting any handsets to be available in time for the Olympics but USB dongles for laptops and some portable media players will be ready and around 37 cities will have transmitters up and running,’ said Ironi. ‘We have seven customer design wins, two with non-Chinese companies, and the rest from local manufacturers.’

In the past Siano has said that it is working with the aigo brand of MP3 players towards a STiMi device in China, and this week added Panda Electronics Group to its list of design wins. “The bill of materials of a CMMB device using Siano’s SMS1180 is surprisingly low, and designing products with it is simple and straightforward on the one hand, while exhibiting top-notch receiver performance on the other, ” said Mu Zhongping, vice chief engineer from Panda Electronics Group, in a statement.

There are multiple local players that are addressing the CMMB STiMi chip market but few who have announced and shown product, two being Innofidei and Telepath Technologies, which was launched by one of the designers of the CMMB technology. Telepath said that it had worked with Infineon and Taiwanese fabricator SMIC and was first to announced a device back in December, while Innofidei only launched its chip last week, and says that it has been working with NeoMagic, a video chip designer working in California, India and Israel.

While we don’t expect the Chinese mobile TV market to take off overnight, it will grow from around 50,000 devices during 2008 to around 26m by the end of 2012, according to Rethink Research projections in our recent report ‘In Search of a Mobile TV Business Model’. However this rate might improve dramatically. So far SARFT has said that the TV service would be free for one year, and that after that it will operate a hybrid business model with some free channels and some paid. If China is modelled based on a paid uptake rate, the 26m number is reached, but if modelled on the free-to-air uptake rates of Korea and Japan, then Chinese shipments could move it rapidly into the global lead in mobile TV.

Failure to get into these markets early may see the existing mobile TV chip vendors, such as DiBcom (which says it has licensed some of its intellectual property in China but not built its own chip), missing out on one of the world’s largest chip markets. Similarly for major players like Texas Instruments, Broadcom, Samsung, NXP and Infineon, which have only partnered in this sector. These may rapidly need to acquire local partners in order to secure a piece of the Chinese mobile TV base, since all the major design wins will be established. The rush will be even bigger if the same Chinese technology begins winning outside market share, for example in parts of Africa, where we know it is being considered for mobile TV services, and potentially in India.

To be shut out of all three of those developing markets in mobile TV might be a considerable blow to any of the major chipmakers. “We estimate a very large market for mobile TV in China,” said Ironi. “We expect adoption rates similar to those seen in Japan and Korea, but with much higher nominal volumes, given the larger population of China and its popularity for multimedia enabled devices.’

Adopting his logic, Japan took around two years to get mobile TV devices to about 12% of handset owners. The same type of uptake for China would be closer to 72m in just a couple of years, but we don’t think this can happen. This would only be possible if the satellite were successfully launched on time and the footprint for mobile TV went from the 90m-plus population in reach now (but they do not all have mobile phones) to virtually every handset in the country (over 600m). A more realistic outcome would be that roughly the same number of handset owners will be in reach in China during the first few years of the service as there were in Japan (around 100m) and that uptake rates at best, if there is a change of direction to the free-to-air model, could be identical in number to Japan and then outstrip Japan during 2011 and 2012.

This analysis was originally published in Faultline, Wireless Watch’s sister title, which offers weekly insights into the digital media markets and the economics of the quad play. Please contact Peter White on [email protected] for more details on Faultline, or the latest Rethink reports on mobile TV, which include the most comprehensive forecasts available for worldwide adoption.