DocumentCode :
474568
Title :
DVFS in loop accelerators using BLADES
Author :
Dasika, Ganesh ; Das, Shidhartha ; Fan, Kevin ; Mahlke, Scott ; Bull, David
Author_Institution :
Adv. Comput. Archit. Lab., Univ. of Michigan, Ann Arbor, MI
fYear :
2008
fDate :
8-13 June 2008
Firstpage :
894
Lastpage :
897
Abstract :
Hardware accelerators are common in embedded systems that have high performance requirements but must still operate within stringent energy constraints. To facilitate short time-to-market and reduced non-recurring engineering costs, automatic systems that can rapidly generate hardware bearing both power and performance in mind are extremely attractive. This paper proposes the BLADES (Better-than-worst-case Loop Accelerator Design) system for automatically designing self-tuning hardware accelerators that dynamically select their best operating frequency and voltage based on environmental conditions, silicon variation, and input data characteristics. Errors in operation are detected by Razor flip-flops, and recovery is initiated. The architecture efficiently supports detection, rollback, and recovery to provide a highly adaptable and configurable loop accelerator. The overhead of deploying Razor flip-flops is significantly reduced by automatically chaining primitive computation operations together. Results on a range of loop accelerators show average energy savings of 32% gained by voltage scaling below the nominal supply voltage.
Keywords :
embedded systems; flip-flops; microprocessor chips; BLADES; DVFS; Razor flip-flops; automatic systems; best operating frequency; better-than-worst-case loop accelerator design; configurable loop accelerator; embedded systems; energy constraint; input data characteristics; loop accelerators; self-tuning hardware accelerators; Blades; Costs; Embedded system; Flip-flops; Hardware; Power engineering and energy; Power generation; Systems engineering and theory; Time to market; Voltage; Embedded systems; Frequency scaling; High-level synthesis; Low power; Voltage scaling;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Design Automation Conference, 2008. DAC 2008. 45th ACM/IEEE
Conference_Location :
Anaheim, CA
ISSN :
0738-100X
Print_ISBN :
978-1-60558-115-6
Type :
conf
Filename :
4555946
Link To Document :
بازگشت