A methodology for the design of dynamic accuracy operators by runtime back bias

International audienceMobile and IoT applications must balance increasing processing demands with limited power and cost budgets. Approximate computing achieves this goal leveraging the error tolerance features common in many emerging applications to reducepower consumption. In particular, adequate (i.e., energy/qualityconfigurable) hardware operators are key components in an error tolerant system. Existing implementations of these operators require significant architectural modifications, hence they areoften design-specific and tend to have large overheads compared to accurate units.In this paper, we propose a methodology to design adequate datapath operators in an automatic way, which uses threshold voltage scaling as a knob to dynamically control the power/accuracy tradeoff. The method overcomes the limitations of previoussolutions based on supply voltage scaling, in that it introduces lower overheads and it allows fine-grain regulation of thistradeoff. We demonstrate our approach on a state-of-the-art 28nm FDSOI technology, exploiting the strong effect of backbiasing on threshold voltage. Results show a power consumption reduction of as much as 39% compared to solutions based onlyon supply voltage scaling, at iso-accurac