### Table 3.5 Rutting rates during HVS loading on Section 501RF Phase Repetitions

1994

### Table 3.4 Rutting rates during HVS loading on Section 503RF Phase Repetitions

1994

### Table 2) and a PR quadtree with the bulk-loading algorithm from Section 7 (denoted \PB-100 quot;).

1999

"... In PAGE 39: ...63 5088 5146 8674 Table 1: Details on line segment maps. Table2 summarizes configurations used for loading the PMR quadtree in the experiments. Three of them use dynamic quadtree insertion (i.... In PAGE 39: ...4. Method B-tree buffering Quadtree buffering Sorting BB-L yes (unlimited) no no BB-M yes (1024 nodes) no no BB-S yes (256 nodes) no yes QB-75 limited yes ( 75% B-tree storage utilization) yes QB-100 limited yes ( 100% B-tree storage utilization) yes Table2 : Summary of PMR quadtree loading methods used in experiments. Table 3 shows the execution time for loading PMR quadtrees for the six data sets using the five load- ing methods.... In PAGE 42: ... An additional difference is that in the experiments above, we used 4-byte integers for the coordinate values of the line segments, while SAND uses 8-byte floating point numbers for coordinate values. For this set of experiments, we used the configurations BB-L , BB-S , and QB-100 , described in Table2 . In keeping with the modest buffering in the latter two, we only buffered 128 of the most recently used disk pages for the relation tuples, where each disk page is 4KB in size, while for BB-L we used a buffer size of 512 disk pages.... In PAGE 43: ... We used SAND for these experi- ments, storing the point geometry directly in the index. We compare using the loading methods BB-L , BB-S , and QB-100 in Table2 , in addition to the PR quadtree bulk-loading algorithm described in Section 7 (denoted below by PB-100 ). Figure 24 shows the execution time of building the quadtree, while Figure 25 shows the speedup of QB-100 compared to BB-L and BB-S .... In PAGE 46: ...) Polys (clust.) Rectangles BB-L BB-S QB-100 Figure 27: Execution time for building PMR quadtrees for polygon data set (la- bels of bars denote loading methods from Table2 ). \Polys (unclust.... ..."

Cited by 10

### Table 5 Contribution of Pavement Layers to Surface Rutting Based on MDD Measurements Surface Rutting Contribution (percent) by Layer under Specified Load for Each Section

2003

"... In PAGE 28: ... Based on these measurements, the contribution of the subbase and subgrade layers was estimated to be less than 20 percent of the total rut depth measured at the pavement surface. Table5 summarizes the approximate contributions of each pavement layer to surface rutting with the MDDs for each... ..."

### TABLE I FINE MODEL CAPACITANCES, AND THE CHARACTERISTIC IMPEDANCES FOR THE SEVEN-SECTION CAPACITIVELY LOADED IMPEDANCE TRANSFORMER

2004

Cited by 4

### Table 1. The fine model capacitances, and the characteristic impedances for the two-section capacitively-loaded impedance transformer.

"... In PAGE 8: ... Assume that the fine model is very expensive and is not recommended for direct optimization. The values of the fine model capacitances are given in Table1 . The characteristic impedances are kept fixed at the optimal values given in Table 1.... In PAGE 8: ... The values of the fine model capacitances are given in Table 1. The characteristic impedances are kept fixed at the optimal values given in Table1 . The physical lengths L1 and L2 ofthetwotransmissionlinesareselectedasdesignableparameters.... ..."

### Table 2. A comparison of the two service schemes in Sections 2 and 3 under heavy loadings.

1999

"... In PAGE 17: ...ith the conventional de nition of tra c intensity =s , = 2:0 and 4.0 in the two cases. Numerical results for these two cases are displayed in Table 2. In Table2 the di erences between the two systems are greater than in Table 1, but still not large. The probability of being eventually served and the mean and standard deviation of the conditional time to be served are very close.... In PAGE 18: ... Also note that the performance is quite di erent from the M/M/s/r model without balking or reneging. Then the steady-state number N is close to s + r, which in Table2 would be 60. The probability of being served is about the same, however.... In PAGE 18: ...eneging. Then the steady-state number N is close to s + r, which in Table 2 would be 60. The probability of being served is about the same, however. In the setting of Table2 the blocking is negligible. The high blocking in M/M/s/r is replaced by balking and reneging in these cases.... ..."

Cited by 23

### Table 2. A comparison of the two service schemes in Sections 2 and 3 under heavy loadings.

1999

"... In PAGE 17: ...ith the conventional de nition of tra c intensity =s , = 2:0 and 4.0 in the two cases. Numerical results for these two cases are displayed in Table 2. In Table2 the di erences between the two systems are greater than in Table 1, but still not large. The probability of being eventually served and the mean and standard deviation of the conditional time to be served are very close.... In PAGE 18: ... Also note that the performance is quite di erent from the M/M/s/r model without balking or reneging. Then the steady-state number N is close to s + r, which in Table2 would be 60. The probability of being served is about the same, however.... In PAGE 18: ...eneging. Then the steady-state number N is close to s + r, which in Table 2 would be 60. The probability of being served is about the same, however. In the setting of Table2 the blocking is negligible. The high blocking in M/M/s/r is replaced by balking and reneging in these cases.... ..."

Cited by 23

### Table 2. Normalized speedup and energy savings for micro- processors with load-value predictors for gcc (Section 3 details our normalization process.)

2005

"... In PAGE 6: ...Surprisingly, we observe in Table2 that for gcc none of the op- timizations is energy-efficient despite the fact that no optimiza- tion slowed down the CPU (Figure 3). Note that the normalized speedup and normalized energy savings are derived from our metric described in Section 3 and should not be confused with Figure 3 or Figure 4.... ..."

Cited by 2