A neuro-fuzzy fan speed controller for dynamic management of processor fan power consumption

The progress in silicon process technology have prepared the processors with more amount of cores. The more cores lead to increase the amount of required power. Therefore, the heat will be increased for the shortage of die size. The high temperature can cause degradation in performance, reliability, transistor aging, transition speed, and an increase in leakage current. One of the primitive thermal management technique is to use cooling equipment which can decrease temperature with no performance reduction. Although the high speed of fan reduces the temperature, it also brings the more power consumption. The increase in fan speed causes an increase in power consumption, large noise levels in the system, and a decrease of fan lifetime that may impact reliability. The Neuro-Fuzzy (NF) fan speed controller (NFSC) that we offered decreases fan power consumption while preventing the temperature violation from an expected temperature. NFSC estimates the minimum required fan speed that holds the temperature close to the desired temperature with the aim of saving power. The primarily practical results demonstrate that our suggested model in compared with the traditional fan controller is significantly able to reduce the average of fan power consumption approximately 30% with increasing of average temperature by 9% (4.7°C) compared to the traditional fan controller.