The impact of If-conversion and branch prediction on program execution on the Intel/sup R/ Itanium/sup TM/ processor

2005 ◽  
Author(s):  
Youngsoo Choi ◽  
A. Knies ◽  
L. Gerke ◽  
Tin-Fook Ngai
Keyword(s):  
Branch Prediction ◽  
Program Execution ◽  
The Impact

Characterizing the impact of predicated execution on branch prediction

1994 ◽  
Author(s):  
Scott A. Mahlke ◽  
Richard E. Hank ◽  
Roger A. Bringmann ◽  
John C. Gyllenhaal ◽  
David M. Gallagher ◽  
...  
Keyword(s):  
Branch Prediction ◽  
Predicated Execution ◽  
The Impact

The impact of unresolved branches on branch prediction scheme performance

1994 ◽  
Vol 22 (2) ◽  
pp. 12-21 ◽  
Author(s):  
A. R. Talcott ◽  
W. Yamamoto ◽  
M. J. Serrano ◽  
R. C. Wood ◽  
M. Nemirovsky
Keyword(s):  
Branch Prediction ◽  
Prediction Scheme ◽  
The Impact

ADAPTIVE PIPELINE VOLTAGE SCALING IN HIGH PERFORMANCE MICROPROCESSOR

2010 ◽  
Vol 19 (08) ◽  
pp. 1817-1834
Author(s):  
CHANG-CHING YEH ◽  
KUEI-CHUNG CHANG ◽  
TIEN-FU CHEN ◽  
CHINGWEI YEH
Keyword(s):  
Performance Improvement ◽  
High Performance ◽  
Supply Voltage ◽  
Branch Prediction ◽  
Voltage Scaling ◽  
Program Execution ◽  
Superscalar Processor ◽  
Great Performance ◽  
Branch Misprediction ◽  
A Performance

Deep pipeline has traditionally been widely used in high performance microprocessor. To allow continuous program execution, branch prediction provides a necessary method of speculatively executing instructions without compromising performance. However, branch misprediction penalty significantly impacts the performance of the deep pipeline processor. This study presents a new Adaptive Pipeline Voltage Scaling (APVS) technique to reduce branch misprediction penalty. For a likely mispredicted branch entering the processor, APVS begins increasing voltage and merging deep pipeline whereby shorter pipeline length permits less branch misprediction penalty. Once the branch is resolved, the merged stages are split and the supply voltage is reduced again. With dedicated shorter pipeline length within each branch misprediction, APVS achieves great performance improvement. The evaluation of APVS in a 13-stage superscalar processor with benchmarks from SPEC2000 applications shows a performance improvement (between 3–12%, average 8%) over baseline processor that does not exploit APVS.

Sign in / Sign up

Export Citation Format

Share Document