The mobile communications industry has lived on the edge of power consumption problems, throughout its existence, and has almost come to take for granted the ability of Moore’s Law to shrink circuitry and shrink the power consumption, allowing each generation of phone to add functions. Unfortunately as consumers have also taken to slimmer and smaller phones the industry now faces a power density problem as well as one of power consumption, Neuvo said at the International Electronics Forum, organized here by Future Horizons.
Smart phones are leveraging multicore architectures to get the performance they need, Neuvo observed but said fuel cells are not really any use as a mobile phone cannot consume more than 3 W to 6 W depending on its size, as it would then require active cooling. As active cooling is not an option better system and IC design must be the ways forward, as they have been before, Neuvo said. The challenges include developing processors that are an order of magnitude more power efficient than today’s models, dissipating 100 microwatts per MIPS rather than today’s 1 milliwatt per MIPS. Better system-level and silicon-level design needs to focus on power consumption and heat dissipation, leakage current, process variations, interconnection losses and capacitances, he said.
At the same time RF integration is facing a discontinuity as the relative silicon are devoted to analog and digital blocks has turned back. In the earliest days mobile communications was entirely analog. In the second generation digital circuitry and advances enabled by Moore’s Law came to dominate the chipsets. But as the digital blocks have shrunk the RF circuitry is becoming large compared to digital. Neuvo said he expects this to be exacerbated further if and when software defined radio (SDR) is deployed.
“The vision of SDR is to have a generic digital hardware that can be programmed to any radio standard and in the next ten years when SDR becomes reality, we foresee a radical change in RF technology,” said Neuvo.
Further, the maturing of semiconductor industry will lead to an era of competitive re-aggregation, Neuvo predicted. This increases the importance of cooperation between research organizations, academia and industry. Neuvo said that within the European Union in the field of embedded computer systems, the Artemis collaborative research program is a good example of this new cooperation bringing together players from different industries and different parts of the value chain.
Related articles:
Capacitive sensors can replace mechanical switches for touch control
Improve the efficiency of mobile device backlighting
Nanogenerator takes a charge from motion