Exploiting Prediction to Reduce Power on Buses

Victor Wen

EECS Department
University of California, Berkeley
Technical Report No. UCB/CSD-03-1294
November 2003

http://www.eecs.berkeley.edu/Pubs/TechRpts/2003/CSD-03-1294.pdf

We explore the possibility of reducing energy consumed by on-chip buses via stateful and stateless coding techniques. We explore the design of a number of simple coding schemes and simulate them using a modified SimpleScalar simulator and SPEC benchmarks. We show an average of 36% savings in transitions on internal buses such as the reorder buffer and register file. To quantify actual power savings, we design a simple dictionary based encoder/decoder circuit in a 0.13-micron process, extract it as a netlist, and simulate its behavior under SPICE. Utilizing a realistic wire model with repeaters, we show that we can break even at median length scales of less than 11.5mm at 0.13 microns and project a break-even point of 2.7mm for a larger design at 0.07 microns.


BibTeX citation:

@techreport{Wen:CSD-03-1294,
    Author = {Wen, Victor},
    Title = {Exploiting Prediction to Reduce Power on Buses},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {2003},
    Month = {Nov},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/2003/6365.html},
    Number = {UCB/CSD-03-1294},
    Abstract = {We explore the possibility of reducing energy consumed by on-chip buses via stateful and stateless coding techniques. We explore the design of a number of simple coding schemes and simulate them using a modified SimpleScalar simulator and SPEC benchmarks. We show an average of 36% savings in transitions on internal buses such as the reorder buffer and register file. To quantify actual power savings, we design a simple dictionary based encoder/decoder circuit in a 0.13-micron process, extract it as a netlist, and simulate its behavior under SPICE. Utilizing a realistic wire model with repeaters, we show that we can break even at median length scales of less than 11.5mm at 0.13 microns and project a break-even point of 2.7mm for a larger design at 0.07 microns.}
}

EndNote citation:

%0 Report
%A Wen, Victor
%T Exploiting Prediction to Reduce Power on Buses
%I EECS Department, University of California, Berkeley
%D 2003
%@ UCB/CSD-03-1294
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/2003/6365.html
%F Wen:CSD-03-1294