Today at Mobile World Congress, mobile processor juggernaut Texas Instrument held a press event to offer some additional details about its OMAP 4 and 5 processor lines. Greg Delagi, the company’s senior vice president and general manager, gave the keynote — which took place right on the showroom floor. He emphatically talked about the need to push the capabilities of “smart devices” forward while being mindful of the constraints battery technology puts on the industry. Both RIM and LG were brought on stage — the PlayBook and Optimus 3D both use the OMAP 4 platform — and they too heralded the headway TI is making with its system on a chip (SoC) OMAP processors. The company even has technology in place that will bring cost effective, power efficient gesture recognition and Pico DLP technology to a wider array of smartphones and smart devices; it works just like the Kinect. We have some video of the motion-based gesture recognition, in action running on prototype hardware after the break, along with some of Mr. Delagi’s thoughts on why OMAP is such an effective mobile processor.
Delagi was kind enough to give us 10 minutes of his time to talk about the OMAP platform and, more specifically, how it works in the mobile device arena. Mr. Delagi explained that his company’s chip is superior not because of raw processor clock speed — a common myth — but because of exactly how the chip is designed. OMAP 4 and OMAP 5 processors have six separate engines at their disposal. There is the general purpose processor, which is used for most tasks; the graphics processor, used for image display; a programmable DSP processor, which allows manufactures to keep a phone’s core processing capabilities up to the latest codecs (say H.265 came out, this codec could be flashed to the DSP); a display processor for handling the screen’s needs; an imaging and video accelerator, for doing exactly what its name suggests; and the ABE, or audio back-end.
TI uses some extremely advanced system management to power these systems up and down as your phone demands, to try and efficiently utilize every last bit of juice from your battery. For example, the IVA uses 273 milliwatts when running at full throttle. Delagi also noted that the motion-based gesture recognition technology — which you’ll see in the video below — could be added to LG’s Optimus 3D handset with almost zero impact on battery life and/or added cost to the manufacturer. Mobile processor technology is extremely complex, so it was great to hear someone talk so passionately and eloquently about his company’s solution.
A note about this gesture recognition technology: it has a range of 0-8 meters; it can be included in TVs, mobile phones, gaming systems, etc.; it uses no special sensors, just a camera; it has facial recognition capabilities (the example given was if you were trying to change the channel of your TV using gestures but were on the couch with others); and any device with an OMAP 4 or 5 processor can harness this built-in feature.