Filters








560 Hits in 2.8 sec

Scheduling with Predictions and the Price of Misprediction [article]

Michael Mitzenmacher
2019 arXiv   pre-print
As part of our analysis, we suggest the framework of the "price of misprediction," which offers a measure of the cost of using predicted information.  ...  Our main result is the derivation, under natural assumptions, of formulae for the performance of several strategies for queueing systems that use predictions for service times in order to schedule jobs  ...  expressions for the price of misprediction.  ... 
arXiv:1902.00732v2 fatcat:47brl5qacvdqfohpdjgxkyjwcu

Scheduling with Predictions and the Price of Misprediction

Michael Mitzenmacher, Michael Wagner
2020 Innovations in Theoretical Computer Science  
As part of our analysis, we suggest the framework of the "price of misprediction," which offers a measure of the cost of using predicted information.  ...  Our main result is the derivation, under natural assumptions, of formulae for the performance of several strategies for queueing systems that use predictions for service times in order to schedule jobs  ...  expressions for the price of misprediction.  ... 
doi:10.4230/lipics.itcs.2020.14 dblp:conf/innovations/Mitzenmacher20 fatcat:mvcphv4r7rfvtpx773so6zlqum

Half-price architecture

Ilhyun Kim, Mikko H. Lipasti
2003 SIGARCH Computer Architecture News  
We show that a pipeline that optimizes scheduling and register access for a single operand achieves nearly the same performance as an ideal base machine that fully handles two operands, with 2.2% (worst  ...  Two techniques are proposed and evaluated: one for the wakeup logic is sequential wakeup, which decouples half of the tag matching logic from the wakeup bus to reduce the load capacitance of the bus.  ...  Acknowledgements This work was supported in part by the National Science Foundation with grants CCR-0073440, CCR-0083126, EIA-0103670 and CCR-0133437, and generous financial support and equipment donations  ... 
doi:10.1145/871656.859623 fatcat:ewgcpoergrgijgup3viu7zkksq

Half-price architecture

Ilhyun Kim, Mikko H. Lipasti
2003 Proceedings of the 30th annual international symposium on Computer architecture - ISCA '03  
We show that a pipeline that optimizes scheduling and register access for a single operand achieves nearly the same performance as an ideal base machine that fully handles two operands, with 2.2% (worst  ...  Two techniques are proposed and evaluated: one for the wakeup logic is sequential wakeup, which decouples half of the tag matching logic from the wakeup bus to reduce the load capacitance of the bus.  ...  Acknowledgements This work was supported in part by the National Science Foundation with grants CCR-0073440, CCR-0083126, EIA-0103670 and CCR-0133437, and generous financial support and equipment donations  ... 
doi:10.1145/859618.859623 fatcat:gyuky6ao5zahndqdpamuerka7m

Half-price architecture

Ilhyun Kim, Mikko H. Lipasti
2003 Proceedings of the 30th annual international symposium on Computer architecture - ISCA '03  
We show that a pipeline that optimizes scheduling and register access for a single operand achieves nearly the same performance as an ideal base machine that fully handles two operands, with 2.2% (worst  ...  Two techniques are proposed and evaluated: one for the wakeup logic is sequential wakeup, which decouples half of the tag matching logic from the wakeup bus to reduce the load capacitance of the bus.  ...  Acknowledgements This work was supported in part by the National Science Foundation with grants CCR-0073440, CCR-0083126, EIA-0103670 and CCR-0133437, and generous financial support and equipment donations  ... 
doi:10.1145/859622.859623 fatcat:3sj5pmh5bzhazi7b5ygbkdqbum

A trace cache microarchitecture and evaluation

E. Rotenberg, S. Bennett, J.E. Smith
1999 IEEE transactions on computers  
The microarchitecture provides high instruction fetch bandwidth with low latency by explicitly sequencing through the program at the higher level of traces, both in terms of (1) control flow prediction  ...  However, for one benchmark whose performance is limited by branch mispredictions, the performance gain is due almost entirely to improved prediction accuracy.  ...  This work was supported in part by NSF Grants MIP-9505853 and MIP-9307830 and by the U.S. Army Intelligence Center and Fort Huachuca under Contract DABT63-95-C-0127 and ARPA order no. D346.  ... 
doi:10.1109/12.752652 fatcat:5nrm3ihc5rcpzlqop3dkpmkcjq

Pricing Python parallelism: a dynamic language cost model for heterogeneous platforms

Dejice Jacob, Phil Trinder, Jeremy Singer
2020 Proceedings of the 16th ACM SIGPLAN International Symposium on Dynamic Languages  
The results show small misprediction intervals and a mean slowdown of just 13.6%, relative to the optimal (oracular) offload strategy.  ...  Predicting execution time in a language like Python is extremely challenging, since aspects like the element types, size of the iteration space, and amenability to parallelization can only be determined  ...  Acknowledgments The authors would like to thank Alexandre Bergel for his friendly and constructive shepherding of this paper. We also thank the anonymous reviewers for their helpful suggestions.  ... 
doi:10.1145/3426422.3426979 fatcat:ex2h76pov5dgtiysm7ap4rn5de

Application Resource Demand Phase Analysis and Prediction in Support of Dynamic Resource Provisioning

Jian Zhang, Mazin Yousif, Robert Carpenter, Renato J. Figueiredo
2007 Fourth International Conference on Autonomic Computing (ICAC'07)  
The process takes into consideration that application's resource consumption patterns, pricing schedules defined by the resource provider and penalties associated with Service-Level Agreement (SLA) violations  ...  The phase profile learned from historical runs is used to classify and predict phase behavior.  ...  the pricing schedule of the resource owner, and the penalty associated with SLA violations caused by mispredictions. 2.  ... 
doi:10.1109/icac.2007.7 dblp:conf/icac/ZhangYCF07 fatcat:aczxqiexung2rg7mar7vapf3ca

A system level perspective on branch architecture performance

B. Calder, D. Grunwald, J. Emer
1995 Proceedings of the 28th Annual International Symposium on Microarchitecture  
We also show that the most commonly used branch performance metrics, branch misprediction rates and the branch execution penalty, are highly correlated with program performance and are suitable metrics  ...  Fetch prediction is the process of determining the next instruction to request from the memory subsystem. Branch prediction is the process of predicting the likely out-come of branch instructions.  ...  Acknowledgments We could not have completed this study without significant equipment and software support from Digital Equipment Corporation, and the cooperation and assistance of the many people who have  ... 
doi:10.1109/micro.1995.476827 dblp:conf/micro/CalderGE95 fatcat:akpgig4eevdmhj7hj5akmej564

Optimal Liability when Consumers Mispredict Product Usage

Andrzej Baniak, Peter Grajzl
2016 American Law and Economics Review  
Since the expected consumer harm directly depends on the level of product usage, the consequences of consumers' mispredictions vary with the prevailing liability regime.  ...  Our analysis clarifies why and how the choice of the socially optimal legal regime depends on the distribution of consumers based on the direction and extent of their mispredictions.  ...  Instead, when choosing the level of precaution and determining the price in period 1, the firms must predict the consumers' future level of activity.  ... 
doi:10.1093/aler/ahw017 fatcat:dti4iiwfp5gj3lmk6qatxhneye

Optimal Liability When Consumers Mispredict Product Usage

Andrzej Baniak, Peter Grajzl
2016 Social Science Research Network  
Since the expected consumer harm directly depends on the level of product usage, the consequences of consumers' mispredictions vary with the prevailing liability regime.  ...  Our analysis clarifies why and how the choice of the socially optimal legal regime depends on the distribution of consumers based on the direction and extent of their mispredictions.  ...  Instead, when choosing the level of precaution and determining the price in period 1, the firms must predict the consumers' future level of activity.  ... 
doi:10.2139/ssrn.2763632 fatcat:e3e7efdr7rh2jhsngfpwfzrrr4

Instruction pre-processing in trace processors

Q. Jacobson, J.E. Smith
1999 Proceedings Fifth International Symposium on High-Performance Computer Architecture  
A processor that uses traces as the unit of sequencing and execution achieves high instruction fetch rates and can support very wide-issue execution engines.  ...  Traces are "pre-processed" to optimize the instructions for execution together. We propose three specific optimizations: instruction scheduling, constant propagation, and instruction collapsing.  ...  Next-trace prediction [4] implicitly performs branch prediction and branch target prediction with sufficient bandwidth to take advantage of the potentially high fetch rate of the trace cache.  ... 
doi:10.1109/hpca.1999.744347 dblp:conf/hpca/JacobsonS99 fatcat:igbwi5hnczgtzm5jm4tayutqwy

Speculation in elastic systems

Marc Galceran-Oms, Jordi Cortadella, Mike Kishinevsky
2009 Proceedings of the 46th Annual Design Automation Conference on ZZZ - DAC '09  
Speculation is a well-known technique for increasing parallelism of the microprocessor pipelines and hence their performance.  ...  The technique is based on applying provably correct transformations. The benefits of speculation are illustrated with two examples in which these transformations are systematically applied.  ...  Besides, mispredictions can be detected because of back-pressure on the predicted channel. For correctness of behavior a scheduler should avoid potential scheduling deadlocks.  ... 
doi:10.1145/1629911.1629989 dblp:conf/dac/OmsCK09 fatcat:mbpjcxrd6bhtnbftqojkkftpjy

EOLE

Arthur Perais, André Seznec
2014 SIGARCH Computer Architecture News  
As a result, recovering from mispredictions via selective replay can be avoided and a much simpler mechanism -pipeline squashing -can be used, while the out-of-order engine remains mostly unmodified.  ...  Nonetheless, recent work in the field of VP has shown that given an efficient confidence estimation mechanism, prediction validation could be removed from the out-of-order engine and delayed until commit  ...  Acknowledgment This work was partially supported by the European Research Council Advanced Grant DAL No. 267175  ... 
doi:10.1145/2678373.2665742 fatcat:4tapabt7pffgvm2j5hcxdgabti

EOLE: Paving the way for an effective implementation of value prediction

Arthur Perais, Andre Seznec
2014 2014 ACM/IEEE 41st International Symposium on Computer Architecture (ISCA)  
As a result, recovering from mispredictions via selective replay can be avoided and a much simpler mechanism -pipeline squashing -can be used, while the out-of-order engine remains mostly unmodified.  ...  Nonetheless, recent work in the field of VP has shown that given an efficient confidence estimation mechanism, prediction validation could be removed from the out-of-order engine and delayed until commit  ...  Acknowledgment This work was partially supported by the European Research Council Advanced Grant DAL No. 267175  ... 
doi:10.1109/isca.2014.6853205 dblp:conf/isca/PeraisS14 fatcat:s7ly2chn2jgpta6yh3425bzoaq
« Previous Showing results 1 — 15 out of 560 results