### Table 5: Performance Comparison between [12] and [14]. System Processing Elements Performance CLK

"... In PAGE 4: ... That implies writing sev- eral nucleotides in a single bus write- cycle. The results from Table5 show that their Virtex2 implementation is the fastest in terms of CUPS (Cell Updates Per Second). In [13] a concept to accelerate the S-W algorithm on the bases linear systolic array is demonstrated.... In PAGE 5: ... The implementation was successfully veri ed using on Pilchard (a recon gurable computing platform), which provides a 133 MHz, 64-bit wide memory mapped bus to the FPGA. Table5 gives a performance comparison between [12] and [14], where the performance is measured in billion cell updates per second (BCUPS). It is obvious from Table 5, that the Virtex 2 implementation in [12] outperforms other implementations in terms of BCUPS.... ..."

### Table 2. Questions for the participants and results. Note that participants on questions 10 and 12 to 14 significantly favored the optical system.

### Table 2: Bounds (10), (12), (14), (24), and (26) for Examples 4, 5, 6, and 14.

"... In PAGE 20: ... A Comparison In this appendix we tabulate a brief comparison of some of the upper and lower bounds derived in the body of this paper. Table2 reports bounds (10), (12), (14), (15), (24), and (26) for Examples 4, 5, 6, and the following engineering application. Example 14 The simpli ed, linearized model of a two-dimensional, three-link mobile manipulator derived in [25] is a linear, time invariant descriptor control system E _ x = Ax + Bu y = Cx: The explicit data listed in [25] are E = 2 4 I3 0 0 0 M0 0 0 0 0 3 5 ; A = 2 4 0 I3 0 ?K0 ?D0 FT 0 F0 0 0 3 5 ; B = 2 4 0 S0 0 3 5 ; and C = 2 4 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 3 5 ; where M0 = 2 4 18:7532 ?7:94493 7:94494 ?7:94493 31:8182 ?26:8182 7:94494 ?26:8182 26:8182 3 5 ; D0 = 2 4 ?1:52143 ?1:55168 1:55168 3:22064 3:28467 ?3:28467 ?3:22064 ?3:28467 3:28467 3 5 ; K0 = 2 4 67:4894 69:2393 ?69:2393 69:8124 1:68624 ?1:68617 ?69:8123 ?1:68617 ?68:2707 3 5 ; S0 = 2 4 ?0:216598 ?:033806 0:554659 0:458506 ?0:845154 0:386648 ?:458506 0:845153 0:613353 3 5 ; F0 = 1 0 0 0 0 1 : For Table 2 we applied the bounds to (A; E) for the open-loop pencil A ? E.... In PAGE 20: ... Table 2 reports bounds (10), (12), (14), (15), (24), and (26) for Examples 4, 5, 6, and the following engineering application. Example 14 The simpli ed, linearized model of a two-dimensional, three-link mobile manipulator derived in [25] is a linear, time invariant descriptor control system E _ x = Ax + Bu y = Cx: The explicit data listed in [25] are E = 2 4 I3 0 0 0 M0 0 0 0 0 3 5 ; A = 2 4 0 I3 0 ?K0 ?D0 FT 0 F0 0 0 3 5 ; B = 2 4 0 S0 0 3 5 ; and C = 2 4 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 1 3 5 ; where M0 = 2 4 18:7532 ?7:94493 7:94494 ?7:94493 31:8182 ?26:8182 7:94494 ?26:8182 26:8182 3 5 ; D0 = 2 4 ?1:52143 ?1:55168 1:55168 3:22064 3:28467 ?3:28467 ?3:22064 ?3:28467 3:28467 3 5 ; K0 = 2 4 67:4894 69:2393 ?69:2393 69:8124 1:68624 ?1:68617 ?69:8123 ?1:68617 ?68:2707 3 5 ; S0 = 2 4 ?0:216598 ?:033806 0:554659 0:458506 ?0:845154 0:386648 ?:458506 0:845153 0:613353 3 5 ; F0 = 1 0 0 0 0 1 : For Table2 we applied the bounds to (A; E) for the open-loop pencil A ? E.... In PAGE 21: ...Table 2: Bounds (10), (12), (14), (24), and (26) for Examples 4, 5, 6, and 14. In Table2 , the bound (14) requires a choice of Q and Z in the generalized Schur decomposition. For Examples 4, 5, and 6 we used Q = Z = I, because in these three examples E and A are already upper triangular.... ..."

### Table 2. Results of processor-scalability predictions for 12, 14, and 16 node cases in the EP benchmark.

in ScoPred—Scalable User-Directed Performance Prediction Using Complexity Modeling and Historical Data

"... In PAGE 7: ... In the first experiment, we test our system for processor scalability, using data from the A and B columns. Table2 shows the results of using our system to model either the A or B column, i.e.... In PAGE 8: ...Table2 , the system assigns a value of 1/3996971.35 for B1 and a value of 0.... In PAGE 8: ...72 P) + 0.085 We observe in Table2 that our predictions tend to be a bit low, though all predictions are well within a 10% deviation from the mean of the observations. While our predictions for Class A come very close, the mean of the observations does not fall within the 95% confidence or prediction intervals for the 14 and 16 node predictions.... ..."

### Table 2. Results of processor-scalability predictions for 12, 14, and 16 node cases in the EP benchmark.

in ScoPred—Scalable User-Directed Performance Prediction Using Complexity Modeling and Historical Data

"... In PAGE 7: ... In the first experiment, we test our system for processor scalability, using data from the A and B columns. Table2 shows the results of using our system to model either the A or B column, i.e.... In PAGE 8: ...Table2 , the system assigns a value of 1/3996971.35 for B1 and a value of 0.... In PAGE 8: ...72 P) + 0.085 We observe in Table2 that our predictions tend to be a bit low, though all predictions are well within a 10% deviation from the mean of the observations. While our predictions for Class A come very close, the mean of the observations does not fall within the 95% confidence or prediction intervals for the 14 and 16 node predictions.... ..."

### Table 1: Load Balancing Statistics [18] Shasha D. and Goodman N. Concurrent Search Tree Algorithms, ACM Transactions on Database Systems, 13(1), 1988, pp. 53-90. [19] Weihl E. W. and Wang P. Multi-version Memory: Software cache Management for Concurrent B- Trees, Proceedings of the 2nd IEEE Symposium on Parallel and Distributed Processing, 1990, pp. 650-655. [20] Yen I. and Bastani F. Hash Table in Massively Parallel Systems, Proceedings of the 1992 Interna- tional Conferences on Computer Languages, April 20-23, 1992, pp. 660-664.

1992

"... In PAGE 16: ... With hot spots the variation is much greater, indicating the nice e ect load balancing has for smoothing the variation and reducing the gradient. Finally Table1 shows the calculated average number of moves made by a node in the entire system, with and without hot spots and with and without load balancing, and the normalized variation of the capacity at each processor from the mean. The table shows that the load balancing reduces the coe cient of variation at the cost of a very small increase in the average moves in the system, indicating that load balancing is e ective with low overhead.... ..."

Cited by 4

### Table 5: Multiple levels of cryptanalytic difficulty in RC5 [31] Rounds 4 6 8 10 12 14 16

2003

"... In PAGE 4: ....1.5 Energy Consumption Versus Security Trade-offs If different security levels can be provided in a cryptographic al- gorithm, each with associated energy consumption characteristics, a security protocol can be adapted to a level of security commensurate with the current state of the battery of the system. Table5 identifies different security levels for the RC5 cipher, obtained by changing the number of rounds used in the cipher, for a given key and block size (128 bits). Each entry indicates the data (number of attempts) needed for a successful attack against RC5 using differential and linear crypt- analysis techniques.... ..."

Cited by 22