Texas Instruments on Tuesday widened the web of development assistance for 2.5G and 3G cellular handsets based on its OMAP technology.
The chipmaker announced that, for manufacturers adopting its Open Multimedia Applications Platform technology, it would offer support through a network of labs. TI's Independent OMAP Technology Centers, hosted by companies such as Bsquare, will assist cell phone makers in the design and development of OMAP-based handset hardware and software.
OMAP incorporates digital signal processor (DSP) cores, operating systems and applications all aimed at supporting new uses for cellular phones that take advantage of the greater bandwidth made available by forthcoming 2.5G and 3G services. Those uses are expected to include downloading pictures or short video clips and finding directions on the fly.
TI has already announced US$100 million in funding for OMAP application developers. However, because OMAP is a new technology, handset makers will need to invest time and money in creating new handsets. Designing the hardware and then tuning applications and operating systems to run on it is no small task.
"We want to make life easy for the developers," said Gilles Delfassy, vice president and manager of TI's worldwide wireless communications business.
The advent of 2.5G and 3G networks will lead to an explosion of devices and applications, he predicted. But not all of them will be successful, he said.
"I guess that many of them will be total failures, but there will be a couple of them that will emerge," he said.
Aiming to increase the chance of success for its customers, TI has established its first two Independent OMAP Technology Centers at Bsquare of San Jose, California, and Productivity Systems of Richardson, Texas.
The centers are also a way for TI to get a leg up on the competition. As the No 1 supplier of DSP chips, which refine signals for devices such as cell phones, TI is the incumbent in the market, but that doesn't mean all handset makers will adopt OMAP. Competitors such as Intel have their own wares that they'll be pitching to manufacturers. Intel offers a Personal Internet Client Architecture, a kit for building cellular phones and personal digital assistants.
It's likely that whoever grabs the most design wins and garners the most developer support will win the day. TI has nabbed some of the better-known names, such as Nokia, for OMAP. But Intel has also gained backers for its architecture, including 26 Asian manufacturers that have pledged to support it.
TI and Intel are coming at wireless technology from two different directions.
Dallas-based TI believes it's the chatter that matters--that wireless devices exist primarily to allow people to communicate, rather than compute. The network, and by extension the phone, is what makes this possible, so communications chips will be more important than microprocessors, the company says.
Intel, by contrast, has emphasized computing power over communications as the most important aspect of 2.5G and 3G phones, with the advent of new data-oriented applications.
"Real-time algorithms"--what voice is to a cell phone--"will continue to be of primary importance," TI's Delfassy said. Intel has "never been involved in wireless communications".
The difference between the two disciplines comes down to an emphasis on efficiency, according to Delfassy. TI seeks to provide communications chips that complete as much work as possible in a minimum of cycles, therefore reducing the amount of computing power necessary. Completing a task such as voice compression in fewer steps saves power, increasing the amount of time person can chat on a cellular phone.