Title :
Reducing leakage power in wearable medical devices using memory nap controller
Author :
Olorode, Oluleye ; Nourani, M.
Author_Institution :
Dept. of Electr. Eng., Univ. of Texas at Dallas Richardson, Richardson, TX, USA
Abstract :
The increasing demand for wearable medical devices, coupled with their short battery life, has led to a renewed interest in low power processors. A significant proportion of the total power consumed by these medical device processors, occur in the memory systems, and therefore the focus of sour work. In this paper, we propose a technique that reduces the static power consumption in caches with no side effect on processor perforemance. Our proposed architecture achieves this power saving by deterministically lowering the power state of cache lines that are guaranteed not to be accessed in immediate future cycles. We simulated our architecture across different cache configurations, using widely known CAD tools, and observed up to 92% reduction in static power consumption on SPEC2006 benchmarks with no performance penalties and minimal hardware overhead.
Keywords :
CAD; biomedical electronics; biomedical telemetry; body sensor networks; cache storage; deterministic algorithms; low-power electronics; medical control systems; power consumption; telemedicine; CAD tool; SPEC2006 benchmark; cache configuration; cache line power state lowering; cache static power consumption reduction; deterministic power state lowering; leakage power reduction; low power processor; medical device processor; memory nap controller; memory system; minimal hardware overhead; performance penalty; power consumption reduction side effect; power saving; processor architecture; processor perforemance; short battery life; simulation; total power consumption; wearable medical device; Biomedical monitoring; Decoding; Hardware; Hidden Markov models; Medical diagnostic imaging; Power demand; Random access memory; [Theme based] — wearable computing; medical and automotive electronics;
Conference_Titel :
Circuits and Systems Conference (DCAS), 2014 IEEE Dallas
Conference_Location :
Richardson, TX
DOI :
10.1109/DCAS.2014.6965331