The exponential growth of the semiconductor industry raises serious environmental concerns. Accurate life-cycle data are critical for both appropriate product eco-design and life-cycle analysis (LCA) of new electronic applications. Previous semiconductor LCA efforts were based on generic studies dealing with an average component description such as microprocessors, DRAM or Flash memories without taking into account the broad application range of each component. In this paper, we show that accurate life-cycle data can only be obtained with an application-aware approach. We demonstrate this with a life-cycle energy evaluation of microprocessors for five different applications: from high-performance 32 nm CPUs for servers and laptops to low-power 45 nm processors for set-top boxes and smart phones to ultra-low-power 130 nm MCUs for wireless sensors. For each category, we model the energy of the CMOS processing steps for integrated circuit (IC) fabrication as well as the use phase energy demand including both active and stand-by modes. Results show that life-cycle energy varies by a factor 20000× between the high-performance and the ultra-low-power ends.
展开▼
