### Table 4: Cache behavior of isolated benchmarks

2004

"... In PAGE 30: ...behavior (see Table4 ): those with an L2 cache miss rate higher than 1%1 are considered memory bounded (MEM).... In PAGE 85: ...because of the efficacy of offloading for reducing the temperatures . Metric Optimal Design Factors Temperature Primary: Technique Use Offloading instead of Striping Thermal Violations Primary: Technique Use Offloading instead of Striping Performance Primary: Dual Core Design Secondary: Technique Use Symmetric Striping Power Primary: Power to Unused Core Turn Unused Core Off Table4 : Each metric of interest produced a relatively simple best solution. 5.... In PAGE 85: ... The importance of benchmark on many of the dependent variables indicates that a broad range of appli- cations should be evaluated before identifying key factors. Table4 highlights our overall results and raises some inter- esting possibilities. We do not consider there to be any one right answer because different products will use different criteria for selecting a design point.... In PAGE 85: ... Ignoring these monetary and time costs for the moment allows us to focus of which design points work well regard- less of the feasibility of their implementation. Table4 sum- marizes the design points our results indicate should be pur- sued further.... ..."

### Table 2: CPU time in seconds and the number of isolated solutions computed

2004

"... In PAGE 15: ....8GHz CPU with 1GB memory, with varying dimensions. The parameter values for the test problems in numerical experiments are shown in Table 1. Numerical experiments for each problem listed in Table2 were performed six times. Specifying a different seed number when starting PHoM enabled us to generate a set of homotopy functions and then trace the resulting homotopy curves.... In PAGE 17: ... The numerical results included here are taken from one of five successful runs of PHoM. Table2 shows the cpu time and the number of isolated solutions obtained. As mentioned in Section 4, after the module CMPSc traces of all the homotopy curves, the module Verify checks whether there are any pair of initial points resulting in two approximate solutions very near each other, any homotopy curve determined as a divergent one, and any curve tracing aborted when the number of predicted iterations exceeds predItMax.... ..."

Cited by 18

### Table 3 Performance comparison of solvers in VIPRE

"... In PAGE 15: ... 1. The transient was performed using the direct, ADI, and BICGSTAB solver for the inner iteration and results are summarized in Table3 . Parametrics were performed on the convergence criteria and the best results were obtained with 10 4 on the relative infinite l1 norm of both the inner and outer iteration.... In PAGE 15: ... All cases shown here were executed on a single processor of a Sun Ultrasparc II. The total number of outer and inner iterations required during the transient simulation is shown in Table3 , as well as the computational profile for each of the primary modules. Since an iterative solution is most practical for this problem, it is most worthwhile to focus on comparisons of the ADI and BICGSTAB results.... In PAGE 15: ... In fact, the linear solution provided by BICGSTAB with a tolerance of 10 4 is su ciently accurate to make the number of outer iterations nearly the same as the machine precision DIRECT inner solution. Finally, it is worth noting in Table3 that the average number of inner iterations per outer is about 50% lower with BICGSTAB than with ADI, which is a direct result of the superior error reduction per iteration for BICGSTAB. This is a general advantage of preconditioned Krylov methods over stationary iterative methods such as ADI (Saad, 1996).... ..."

### Table 2: Occurences of tasks related to isolated menu items in the Settings menu.

2000

"... In PAGE 7: ... 4. Difference of performance between groups for tasks related to isolated menu items Given that the sounds have been designed to increase the semantic content of menu items, we expect the difference of performances between the groups obtained on the tasks related to isolated items displayed in Table2 , to increase over the experiment. RESULTS Number of keypresses Overall difference between groups Firstly, we have looked at the total number of keypresses used by each participant for each task over the whole experiment This analysis provided us with a primary indication of the differences of performances between the groups.... In PAGE 9: ...task =42). An ANOVA analysis showed a high significance of the regression (F =6:60;p =0:014). These results show that the evolution of the difference of performance between the groups in terms of errors made is largely in favour of Group 1. Difference of performance between groups for tasks related to isolated menu items To measure this difference, we have performed a linear regression through the differences of normalised keypress number for the eleven tasks presented in Table2 . Figure 4 shows the result of the regression analysis.... ..."

Cited by 10

### Table 2: Occurences of tasks related to isolated menu items in the Settings menu.

2000

"... In PAGE 6: ... 4. Difference of performance between groups for tasks related to isolated menu items Given that the sounds have been designed to increase the semantic content of menu items, we expect the difference of performances between the groups obtained on the tasks related to isolated items displayed in Table2 , to increase over the experiment. RESULTS Number of keypresses Overall difference between groups Firstly, we have looked at the total number of keypresses used by each participant for each task over the whole experiment This analysis provided us with a primary indication of the differences of performances between the groups.... In PAGE 8: ...task =42). An ANOVA analysis showed a high significance of the regression (F =6:60;p =0:014). These results show that the evolution of the difference of performance between the groups in terms of errors made is largely in favour of Group 1. Difference of performance between groups for tasks related to isolated menu items To measure this difference, we have performed a linear regression through the differences of normalised keypress number for the eleven tasks presented in Table2 . Figure 4 shows the result of the regression analysis.... ..."

Cited by 10

### Table 2: Occurences of tasks related to isolated menu items in the Settings menu.

2000

"... In PAGE 6: ... 4. Difference of performance between groups for tasks related to isolated menu items Given that the sounds have been designed to increase the semantic content of menu items, we expect the difference of performances between the groups obtained on the tasks related to isolated items displayed in Table2 , to increase over the experiment. RESULTS Number of keypresses Overall difference between groups Firstly, we have looked at the total number of keypresses used by each participant for each task over the whole experiment This analysis provided us with a primary indication of the differences of performances between the groups.... In PAGE 8: ...task = 42). An ANOVA analysis showed a high significance of the regression (F = 6:60; p = 0:014). These results show that the evolution of the difference of performance between the groups in terms of errors made is largely in favour of Group 1. Difference of performance between groups for tasks related to isolated menu items To measure this difference, we have performed a linear regression through the differences of normalised keypress number for the eleven tasks presented in Table2 . Figure 4 shows the result of the regression analysis.... ..."

Cited by 1

### Table 2: Memory allocation results for isolated sharing.

"... In PAGE 13: ...pectively. We allocate memory in the multiple of 4 MB. We use exhaustive search to find the optimal allocation strategy and the computation overhead for the exhaustive search is acceptable for three VMs. Table2 lists the memory allocation results. Overall, the experimental results show that our hypervisor cache- based memory allocation scheme can substantially re- duce the system-wide page miss metric (15% average page miss reduction at 5% isolation constraint and 59% average page miss reduction at 25% isolation constraint).... ..."

### Table 5: Performance of the optimised parser modules when tested in isolation.

2003

"... In PAGE 11: ... All experiments used 10-fold cross-validation, so that the choices in the optimisation procedure were statistically sound. The parameter settings for the three modules that were found lead to the scores listed in Table5 . It can be seen that all three memory-based parser modules outperform the baseline scores for the corresponding learning tasks by wide margins.... In PAGE 11: ... It can be seen that all three memory-based parser modules outperform the baseline scores for the corresponding learning tasks by wide margins. Performance of the parsing cascade The scores in Table5 correspond to mod- ules for which the input instances were constructed from gold-standard corpus data, that is, the PNP finder and the grammatical relation finder have been run on perfect part of speech and chunk information; likewise, the information about PNP chunks fed to the grammatical relation finder was directly extracted from corpus data. While such a setup is acceptable for the heuristic optimisation procedure, the actual performance of the entire shallow parser can only be measured when the three modules are applied in sequence to the input data.... ..."

Cited by 2

### Table 5: Compute time in CPU seconds objects atts AODE NB TAN LBR J48

2002

"... In PAGE 5: ... 6.2 Relative Compute Time Table5 presents the total compute time of each algorithm by data set. These times are the total time to run Weka to complete the cross-validation task and hence include data input as well as learning and classification.... ..."

Cited by 7

### Table 1. Behavior of some queueing schemes according to isolation and sharing Algorithm isolation sharing bandwidth

in Scheduling

"... In PAGE 2: ... All packet scheduling algorithms try to ensure some level of isolation while conserving the bene ts of sharing. Table1 shows the characteristics of some scheduling algorithms with respect to isolation and sharing. 1 IP speci cations [9] include a precedence (i.... ..."