Filters








786 Hits in 5.0 sec

A comparative analysis of schemes for correlated branch prediction

Cliff Young, Nicolas Gloy, Michael D. Smith
1995 Proceedings of the 22nd annual international symposium on Computer architecture - ISCA '95  
We also use the framework to identify the fundamental differences between static and dynamic correlated branch prediction schemes.  ...  We use the framework to show how a static correlated branch prediction scheme increases branch bias and thus improves overall branch prediction accuracy.  ...  We distinguish aliasing from merging: aliasing combines streams from different static branches, while merging combines streams from one static branch.  ... 
doi:10.1145/223982.224438 dblp:conf/isca/YoungGS95 fatcat:h67lwy65pvfcrh4asmfbm3z2ru

A comparative analysis of schemes for correlated branch prediction

Cliff Young, Nicolas Gloy, Michael D. Smith
1995 SIGARCH Computer Architecture News  
We also use the framework to identify the fundamental differences between static and dynamic correlated branch prediction schemes.  ...  We use the framework to show how a static correlated branch prediction scheme increases branch bias and thus improves overall branch prediction accuracy.  ...  We distinguish aliasing from merging: aliasing combines streams from different static branches, while merging combines streams from one static branch.  ... 
doi:10.1145/225830.224438 fatcat:33tf6fu23rcqlprsiru7diemoq

OPTS: increasing branch prediction accuracy under context switch

Moon-Sang Lee, Young-Jae Kang, Joon-Won Lee, Seung-Ryoul Maeng
2002 Microprocessors and microsystems  
Though many dynamic branch predictors have been proposed to obtain high prediction accuracy, they cannot perform as expected under context switches.  ...  In this paper we measure the effect of context switch on branch prediction, and present a new scheme which saves and restores branch predictor table when a context switch occurs.  ...  Acknowledgments This work was supported by National Research Laboratory Program (Grant 4-207) which is funded by Ministry of Science and Technology.  ... 
doi:10.1016/s0141-9331(02)00041-8 fatcat:e6h7v4l3vbesxiqkkufbd6ggp4

Branch Prediction Using Profile Data [chapter]

Alex Ramirez, Josep L. Larriba-Pey, Mateo Valero
2001 Lecture Notes in Computer Science  
The first approach to branch prediction were static predictors, which always predicted the same direction for a given branch.  ...  It is the ability to predict the outcome of a branch which allows the processor to effectively use a large instruction window, and extract a larger amount of ILP.  ...  Recent dynamic prediction schemes try to organize their tables in a clever way to reduce destructive aliasing [7, 12, 16, 24] . We refer to these predictors as de-aliased schemes.  ... 
doi:10.1007/3-540-44681-8_57 fatcat:nif5fumirnflpgymim5h3wpcqe

Trading conflict and capacity aliasing in conditional branch predictors

Pierre Michaud, André Seznec, Richard Uhlig
1997 Proceedings of the 24th annual international symposium on Computer architecture - ISCA '97  
To address this problem, we propose the skewed branch predictor, a multi-bank, tag-less branch predictor, designed specifically to reduce the impact of conflict aliasing.  ...  Through both analytical and simulation models, we show that the skewed branch predictor removes a substantial portion of conflict aliasing by introducing redundancy to the branch-predictor tables.  ...  Although previous branch-prediction research has shown how to reduce compulsory and capacity aliasing, little has been done to reduce conflict aliasing.  ... 
doi:10.1145/264107.264211 dblp:conf/isca/MichaudSU97 fatcat:ahrhri7mg5gulg6xcfe2htk2ey

Trading conflict and capacity aliasing in conditional branch predictors

Pierre Michaud, André Seznec, Richard Uhlig
1997 SIGARCH Computer Architecture News  
To address this problem, we propose the skewed branch predictor, a multi-bank, tag-less branch predictor, designed specifically to reduce the impact of conflict aliasing.  ...  Through both analytical and simulation models, we show that the skewed branch predictor removes a substantial portion of conflict aliasing by introducing redundancy to the branch-predictor tables.  ...  Although previous branch-prediction research has shown how to reduce compulsory and capacity aliasing, little has been done to reduce conflict aliasing.  ... 
doi:10.1145/384286.264211 fatcat:rijsc5dm6ffbxb4pjly2w3l3v4

OS-Aware Branch Prediction: Improving Microprocessor Control Flow Prediction for Operating Systems

Tao Li, Lizy Kurian John, Anand Sivasubramaniam, N. Vijaykrishnan, Juan Rubio
2007 IEEE transactions on computers  
We propose two simple OS-aware control flow prediction techniques to alleviate the destructive impact of user/OS branch interference.  ...  The OS execution affects various architectural states, including the dynamic branch predictions, which are widely used in today's high-performance microprocessor designs to improve performance.  ...  In this paper, we examine an alternative approach, OS-aware branch prediction, dedicated to cost-effectively reducing the impact of the destructive user/OS branch aliasing without introducing additional  ... 
doi:10.1109/tc.2007.250619 fatcat:ga5ntxnsyfddnexb42ueurfowe

Understanding and improving operating system effects in control flow prediction

Tao Li, Lizy Kurian John, Anand Sivasubramaniam, N. Vijaykrishnan, Juan Rubio
2002 SIGARCH Computer Architecture News  
We propose two simple OS-aware control flow prediction techniques to alleviate the destructive impact of user/OS branch interference.  ...  We study the improvement contributed by the OSaware prediction to various branch predictors ranging from simple Gshare to more elegant Agree, Multi-Hybrid and Bi-Mode predictors.  ...  We would like to thank Steven K. Reinhardt (University of Michigan), David Wood (University of Wisconsin) and all anonymous reviewers for their insightful comments and suggestions.  ... 
doi:10.1145/635506.605405 fatcat:qyji6rt3dbd6dhsepy7o7ffeci

Understanding and improving operating system effects in control flow prediction

Tao Li, Lizy Kurian John, Anand Sivasubramaniam, N. Vijaykrishnan, Juan Rubio
2002 ACM SIGOPS Operating Systems Review  
We propose two simple OS-aware control flow prediction techniques to alleviate the destructive impact of user/OS branch interference.  ...  We study the improvement contributed by the OSaware prediction to various branch predictors ranging from simple Gshare to more elegant Agree, Multi-Hybrid and Bi-Mode predictors.  ...  We would like to thank Steven K. Reinhardt (University of Michigan), David Wood (University of Wisconsin) and all anonymous reviewers for their insightful comments and suggestions.  ... 
doi:10.1145/635508.605405 fatcat:wlv4kduowbhg5itb2ft66l4i4y

Understanding and improving operating system effects in control flow prediction

Tao Li, Lizy Kurian John, Anand Sivasubramaniam, N. Vijaykrishnan, Juan Rubio
2002 Tenth international conference on architectural support for programming languages and operating systems on Proceedings of the 10th international conference on architectural support for programming languages and operating systems (ASPLOS-X) - ASPLOS '02  
We propose two simple OS-aware control flow prediction techniques to alleviate the destructive impact of user/OS branch interference.  ...  We study the improvement contributed by the OSaware prediction to various branch predictors ranging from simple Gshare to more elegant Agree, Multi-Hybrid and Bi-Mode predictors.  ...  We would like to thank Steven K. Reinhardt (University of Michigan), David Wood (University of Wisconsin) and all anonymous reviewers for their insightful comments and suggestions.  ... 
doi:10.1145/605402.605405 fatcat:as6yzqw3qngpbogje45hpwddei

Understanding and improving operating system effects in control flow prediction

Tao Li, Lizy Kurian John, Anand Sivasubramaniam, N. Vijaykrishnan, Juan Rubio
2002 SIGPLAN notices  
We propose two simple OS-aware control flow prediction techniques to alleviate the destructive impact of user/OS branch interference.  ...  We study the improvement contributed by the OSaware prediction to various branch predictors ranging from simple Gshare to more elegant Agree, Multi-Hybrid and Bi-Mode predictors.  ...  We would like to thank Steven K. Reinhardt (University of Michigan), David Wood (University of Wisconsin) and all anonymous reviewers for their insightful comments and suggestions.  ... 
doi:10.1145/605432.605405 fatcat:hz7m2rbkcrdnxp7dmldsirb42a

Understanding and improving operating system effects in control flow prediction

Tao Li, Lizy Kurian John, Anand Sivasubramaniam, N. Vijaykrishnan, Juan Rubio
2002 Tenth international conference on architectural support for programming languages and operating systems on Proceedings of the 10th international conference on architectural support for programming languages and operating systems (ASPLOS-X) - ASPLOS '02  
We propose two simple OS-aware control flow prediction techniques to alleviate the destructive impact of user/OS branch interference.  ...  We study the improvement contributed by the OSaware prediction to various branch predictors ranging from simple Gshare to more elegant Agree, Multi-Hybrid and Bi-Mode predictors.  ...  We would like to thank Steven K. Reinhardt (University of Michigan), David Wood (University of Wisconsin) and all anonymous reviewers for their insightful comments and suggestions.  ... 
doi:10.1145/605397.605405 dblp:conf/asplos/LiJSVR02 fatcat:we2cmgp3ubcdfkgvug6hkluriu

An analysis of dynamic branch prediction schemes on system workloads

Nicolas Gloy, Cliff Young, J. Bradley Chen, Michael D. Smith
1996 SIGARCH Computer Architecture News  
Recent studies of dynamic branch prediction schemes rely almost exclusively on user-only simulations to evaluate performance.  ...  By analyzing our own Atom-generated system traces and the system traces from the Instruction Benchmark Suite, we quantify the effects of kernel and user interactions on branch prediction accuracy.  ...  Both of these trends make intuitive sense, since the larger sizes should reduce aliasing within the combined set of user and kernel branches.  ... 
doi:10.1145/232974.232977 fatcat:onaq2iywrrfwbmqejej5gvjkry

An analysis of dynamic branch prediction schemes on system workloads

Nicolas Gloy, Cliff Young, J. Bradley Chen, Michael D. Smith
1996 Proceedings of the 23rd annual international symposium on Computer architecture - ISCA '96  
Recent studies of dynamic branch prediction schemes rely almost exclusively on user-only simulations to evaluate performance.  ...  By analyzing our own Atom-generated system traces and the system traces from the Instruction Benchmark Suite, we quantify the effects of kernel and user interactions on branch prediction accuracy.  ...  Both of these trends make intuitive sense, since the larger sizes should reduce aliasing within the combined set of user and kernel branches.  ... 
doi:10.1145/232973.232977 dblp:conf/isca/Gloy96 fatcat:k7fq6gu4qffljiyipxlxfplmda

Correlation and aliasing in dynamic branch predictors

Stuart Sechrest, Chih-Chieh Lee, Trevor Mudge
1996 Proceedings of the 23rd annual international symposium on Computer architecture - ISCA '96  
We have examined the sensitivity of these prediction schemes to variation in workload, in resources, and in design and configuration.  ...  Branch Address Prior Branch Outcomes Column Aliasing Row Merging  ...  Conditional Branches (Percent of Total Instructions) Static Condition al Branches # of Static Branches Constituting 90% of Total Dynamic Conditional Branches compress 83947354 11739532  ... 
doi:10.1145/232973.232978 dblp:conf/isca/SechrestLM96 fatcat:pnh34hoqvfbsjjfn6mzrbuftpm
« Previous Showing results 1 — 15 out of 786 results