P. Crowley, J.-L. Baer, On the use of trace sampling for architectural studies of desktop applications, in: Proceedings of the SIGMETRICS'99, Atlanta, GA, 1999. Home/Search Document Details and Download
Summary Related Articles Check |
This paper is cited in the following contexts:
Performance Evaluation 45 (2001) 1--18 - Adaptive Multivariate Regression (Correct)
....are limited to analyzing generally smaller applications since simulations are orders of magnitude slower than actual runs. Due to the large amount of information provided from simulations, statistical reduction methods have been introduced to sift through the enormous amount of data generated [4]. Methods such as micro benchmarking [14] focus on measuring performance of individual code segments while characterizing codes for the number of occurrences of each micro code segment. Iterative performance bounding [11] isolates the performance of singular contributions like compiler and ....
P. Crowley, J.-L. Baer, On the use of trace sampling for architectural studies of desktop applications, in: Proceedings of the SIGMETRICS'99, Atlanta, GA, 1999.
Techniques for Accurate, Accelerated Processor Simulation: .. - Haskins, Jr., Skadron (2002) (Correct)
....execution driven simulator, run in an average of 4 hours, with the worst case finishing in just less than 6.5 hours. 3 Sampling A large body of prior work has explored sampling techniques for computer architecture research. For example, Laha, Patel, and Iyer [19] Crowley and Baer [6]; Martonosi, Gupta, and Anderson [22] Kaplan, Smaragdakis, and Wilson [14] and Elnozahy [9] examine memory reference trace sampling and present new algorithms for trace reduction and compression. Other work has studied analytic models for estimating cache miss rates during the unprimed portion ....
P. Crowley and J.-L. Baer. On the use of trace sampling for architectural studies of desktop applications. In 208--09, May 1999.
Choosing Representative Slices of Program Execution for.. - Lafage, Seznec (2000) (12 citations) (Correct)
....misses at the beginning of the selected slices is the same as in the trace sampling case, at the beginning of the samples. For this reason, methods applied to the trace sampling approach to reduce this non sampling bias [5] such as the use of bigger slices, warm up periods, trace stitching. [1, 7, 6, 28]) may also apply to our approach. Slices selected by our method may be far from the beginning of the program. This necessitates to skip a lot of instructions before the simulation starts as enabled by the direct execution mode of the calvin2 DICE toolset [17] The ultimate use of the method ....
P. Crowley and J. L. Baer. On the use of trace sampling for architectural studies of desktop applications. In Proceedings of the ACM SIGMETRICS International Conference on Measurement and Modeling of Computing Systems (SIGMETRICS-99), volume 27,1 of SIGMETRICS Performance Evaluation Review, pages 208--209, New York, May 1--4 1999. ACM Press.
On the Use of Trace Sampling for Architectural Studies of.. - Crowley, Baer (1998) (3 citations) Self-citation (Crowley Baer) (Correct)
....may be used to assess trends in victim cache performance[2] and the second study uses sampling to assess trends in branch prediction techniques. In addition, we can estimate parameters for analytical cache models with sampled traces as precisely as estimating with whole reference traces (cf. [1] for this part of the study) 2. TRACE SAMPLING In trace sampling, an observation, or sample, is obtained by recording a fixed number, the sample size, of consecutive references from a reference stream. Another fixed number of references are ignored before the next observation is made. The ....
P.J. Crowley and J-L Baer. On the use of trace sampling for architectural studies of desktop applications. Technical Report UW-CSE-98-12-05, University of Washington, 1998.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC