Abstract | ||
---|---|---|
A workload-aware low-power neuromorphic controller for dynamic voltage scaling in VLSI systems is presented. The neuromorphic controller predicts future workload values and preemptively regulates supply voltage based on past workload profile. Our specific contributions include: (1) implementation of a digital and analog version of the controller in 45nm CMOS technology, resulting in 3% performance hit with a power overhead in the range of 10–150 microwatts, (2) higher prediction accuracy compared to a software based OS-governed DVS scheme by 50%, reducing wasted power and improving error margins, (3) digital design has minimal power overhead and is more reconfigurable, while analog design is better suited for nonlinear and complex computational tasks. |
Year | DOI | Venue |
---|---|---|
2010 | 10.1145/1840845.1840896 | ISLPED |
Keywords | Field | DocType |
minimal power overhead,low-power supply voltage controller,power overhead,analog version,digital design,neuromorphic engineering,dvs,dynamic voltage scaling,future workload value,workload-aware neuromorphic design,past workload profile,workload-aware low-power neuromorphic controller,neuromorphic controller,spiking neurons,analog design,radiation detectors,computer architecture | Dynamic voltage scaling,Control theory,Computer science,Workload,Voltage,Neuromorphic engineering,Electronic engineering,CMOS,Voltage controller,Real-time computing,Software | Conference |
ISBN | Citations | PageRank |
978-1-4244-8588-8 | 4 | 0.40 |
References | Authors | |
10 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Saurabh Sinha | 1 | 195 | 21.88 |
Jounghyuk Suh | 2 | 18 | 2.38 |
Bertan Bakkaloglu | 3 | 273 | 48.14 |
Yu Cao | 4 | 329 | 29.78 |