Tools
Sorted by:
Try your query at:
 |
 |
 |
 |
 |
 |
Results 1 - 10
of
111,168
Table 5: The performance of an MPI-chimp based //ELLPACK solvers on a cluster of 16 non- homogeneous workstations (Execution times and speedups).
"... In PAGE 8: ... Total execution time and speed up are presented in each square. Data set 4: Cluster of sixteen non-homogeneous workstations and MPI Chimp The results are indicated in Table5 . The speed up with 16 workstations is about 6 with the largest problem tried.... ..."
Table 7: The performance of an MPI-LAM based //ELLPACK solvers on a cluster of 16 non- homogeneous workstations (Execution times and speedups).
"... In PAGE 10: ...Table 7: The performance of an MPI-LAM based //ELLPACK solvers on a cluster of 16 non- homogeneous workstations (Execution times and speedups). Data set 6: Cluster of sixteen non-homogeneous workstations and MPI-LAM The results are indicated in Table7 . The speed up is maximum 3 for the largest problem tried.... ..."
Table 2: Parameters for injected non-homogeneity and target in MCARM Data
"... In PAGE 40: ... 5 - 36 RTO-EN-SET-063 A synthetic target injected at the look direction and Doppler illustrates that sensitivity of the hybrid algorithm to weak targets. The parameters of the weak target are listed in Table2 . Figure 25(b) compares the output of the two algorithms in the case of a strong non-homogeneity and a weak target.... ..."
Table 1: Comparison of relative performance and CPU time consumption for the di erent algorithms for N=M=30 problems. proaches are compared with the exact BB for N=M=30 non-homogeneous and homogeneous problems. As expected LP and in particular GH bene ts from
Table 3: The performance of MPI-ANL-MSU based //ELLPACK solvers on a cluster 8 non- homogeneous workstations (Execution times and speedups).
"... In PAGE 8: ... We present the total execution time and speed up in every square of the table for each table entry. Data set 2: Cluster of sun LXs, SS2s, and IPCs and MPI-ANL-MSU The results are indicated in Table3 . Total execution time and speed up are presented in each square.... ..."
TABLE IV INCREASED DROPPING/BLOCKING RATIOS FOR THE HOT-SPOT UNDER NON-HOMOGENEOUS TRAFFIC
2000
Cited by 10
Table 1: Attribute Ranges. 1. All the tuples of a relation are homogeneous, i.e., each set-triplet valued attribute of a tuple has the same starting and ending times. Note that TDBMS supports non-homogeneous tuples as well. 2. The instances of each set-triplet valued attribute are continuous, i.e., we assume no null values or discontinuities in the tuples of these relations. 4.3 Generating the TTL Data The test database eval and the relations given in Section 4.2.1 are used to generate the relations which use tuple timestamping where the whole temporal data of a relation is contained within a single relation. These temporal relations are generated by rst unpacking the relations on each set-triplet valued attribute and then intersection slicing each (triplet-valued) attribute by the other (triplet-valued) attribute. This will ensure that each attribute has the same time reference. A droptime operation is then carried out on each temporal attribute except one. Triplet-decomposition on this attribute gives the desired relation. For 14
"... In PAGE 13: ...oating-point values uniformly distributed over the interval [0.0, 1.0), is used to generate random data. The random number is then converted to an attribute value. The range of the attribute values is listed in Table1 . It should be noted that for each of the set-valued and set-triplet valued attributes in TDBMS, two bytes are allocated in each tuple.... In PAGE 18: ...elation, i.e., 50 tuples since employee names are character strings and string comparison is used in the query. More speci cally, employees with names E1, E10-E14, and E100-E143 are selected (Employee names range from E1 to E500 as shown in Table1 ).... ..."
Table 3. Noise estimation (mean standard deviation) using non-homogeneous areas before and after noise reduction, using the Waldkirch dataset.
2002
"... In PAGE 2: ... They require as input an estimate of the noise, which may be known or estimated by the method mentioned above. Table3 shows the noise estimation before... ..."
Cited by 3
Table 1: Columns: (1) PDB code of the molecular complex. (2) The number of atoms of A (without hydrogen atoms). (3) The number of atoms of B (without hydrogen atoms). (4) The rank of the best approximation of the real conformation (the ranking of the algorithm of Norel et al. [NLW+] is given in brackets). (5) The tness value of the best approximation. (6) The tness value of the best geometric t. (7) The \real tness quot; of the natural conformation, determined by the local optimizer. (8) The RMS deviation of the approximation in A. (9) The sequential running time of the docking program of Norel et al. (the time to carry out the scoring) on a SUN SPARC ??. (10) Preprocessing+docking time of our algorithm. The sequential running times have been measured on a SGI POWER CHALLENGE M. The times for the distributed version have been measured on a non-homogeneous workstation cluster. (11) The number of processors (since we are working on a non-homogenous workstation cluster, we computed a performance number for each processor; total performance value = sum of the performance values of the processors).
1995
Cited by 2
Table 3. PSNR values for the two videos with non- homogenous background (10 Eigenimages).
Results 1 - 10
of
111,168
