Results 1 - 10 of 103,365

Table 6. Overhead of replication frameworks. Replication

in Active Replication of Multithreaded Applications
by C. Basile, Zbigniew Kalbarczyk, K. Whisnant, Ravi Iyer 2002
"... In PAGE 17: ... Ratio between execution times and CCBUC4 for a triplicated echo server. For comparing our framework with other replication system/frameworks Table6 provides performance overhead reported for several existing approaches to replication. Table 6 clearly indicates that most replication strategies impose a significant performance overhead.... In PAGE 17: ... For comparing our framework with other replication system/frameworks Table 6 provides performance overhead reported for several existing approaches to replication. Table6 clearly indicates that most replication strategies impose a significant performance overhead. The exception is Eternal with a reported overhead of 3%.... ..."
Cited by 7

Table 6. Overhead of replication frameworks. Replication

in Active replication of multithreaded applications
by Claudio Basile, Zbigniew Kalbarczyk, Keith Whisnant, Ravi Iyer 2002
"... In PAGE 17: ... Ratio between execution times and CC BUC4 for a triplicated echo server. For comparing our framework with other replication system/frameworks Table6 provides performance overhead reported for several existing approaches to replication. Table 6 clearly indicates that most replication strategies impose a significant performance overhead.... In PAGE 17: ... For comparing our framework with other replication system/frameworks Table 6 provides performance overhead reported for several existing approaches to replication. Table6 clearly indicates that most replication strategies impose a significant performance overhead. The exception is Eternal with a reported overhead of 3%.... ..."
Cited by 7

Table 2: Performance comparison between normal OpenSSH and privilege separated OpenSSH. We measure the overhead in login and data transfer time when employ- ing privilege separation. In both cases, privilege separation imposes no significant performance penality.

in Preventing Privilege Escalation
by Niels Provos Markus
"... In PAGE 9: ... The next two tests measure the data transfer time of a 10 MB file filled with random data, with compres- sion enabled, and without compression. The results are shown in Table2 . It is evident that privilege sep- arated OpenSSH does not penalize performance.... ..."

Table 2: Results for two of the tested classes, showing the time and number of tests required by ARTOO and RAND to uncover the first bug in each routine in which they both found at least one bug, and their relative performance. In most cases ARTOO requires significantly less tests to find a bug, but entails a time overhead.

in ARTOO: Adaptive Random Testing for Object-Oriented
by Ilinca Ciupa, Andreas Leitner, Manuel Oriol
"... In PAGE 5: ... 5.2 Results Table2 shows a routine-by-routine comparison of the time-to- first-bug and tests-to-first-bug for both ARTOO and RAND applied to classes ARRAYED LIST and ACTION SEQUENCE. The table shows, for each routine in which both strategies found bugs, the number of tests and time required by each strategy to find a bug in that particular routine.... ..."

Table 2: Comparison of relative performance on SunOS The figures show that in optimised trim, L2imbo outperforms ANSAware RPC in all cases. However, the overhead of the context switch and local communication required in the standard L2imbo configuration does have a significant impact on the figures. Being able to minimise the overheads associated with exchanging messages between the application stubs and the daemon process is clearly an important factor in further improving the performance of L2imbo.

in Experiences of Using Generative Communications To Support Adaptive Mobile Applications
by Adrian Friday, Nigel Davies, Jochen Seitz, Matt Storey, Stephen P. Wade 1999
Cited by 1

Table 4 gives the results of running the straightforward OpenGL application running on a workstation and on the Wall using WireGL, synchronized, as the distribution mechanism. The overhead of distributing data to the display machines causes a significant performance loss compared to a single workstation. The improvement in performance of Myrinet over 100Mb Ethernet is clear.

in Building a large scale, high-resolution, tiled, rear projected, passive stereo display system based on commodity components
by Glenn Bresnahan, Raymond Gasser, Augustinas Abaravichyus, Erik Brisson, Michael Walterman 2003
"... In PAGE 10: ...8 16.7 Table4 : Effect of network type Table 4 gives the results of running the straightforward OpenGL application running on a workstation and on the Wall using WireGL, synchronized, as the distribution mechanism. The overhead of distributing data to the display machines causes a significant performance loss compared to a single workstation.... ..."
Cited by 3

Table 1 reports memory and performance overhead caused by the application of the old and new rules. The increase of memory and performance overhead caused by the application of the new rules depends on the number of j- instructions in the high-level program description. For example, the F benchmark contains a number of j-instructions,

in unknown title
by unknown authors
"... In PAGE 6: ... Table1... In PAGE 7: ...2. Although efficient in the sense of fault coverage, the improved YACCA technique (as shows Table1 ) leads to an increase of memory and performance overhead. As new rules involve duplication of some program operations, performance overhead increases significantly.... ..."
Cited by 1

Table V presents the execution cycles reported by the simulator for this case. As it was expected the cycles spend for ARISE move and execute instructions remain the same. However, they are now accounting for a smaller portion of the total execution time. This is because execution time consumed in reconfiguring the CCU has been significantly increased to the 35% of the total execution time. Increase in the configuration overhead had an impact on the speedup factor which has been reduced from x7.5 to x6.3. However, performance improvements by the exploitation of the ARISE are still significant.

in The ARISE Reconfigurable Instruction Set Extensions Framework
by Nikolaos Vassiliadis, George Theodoridis, Spiridon Nikolaidis

Table 1: Frequency of significant byte patterns

in Very low power pipelines using significance compression
by Ramon Canal, Antonio González, James E. Smith 2000
"... In PAGE 2: ... We performed a study with the Mediabench benchmarks [6] to determine the relative frequency of occurrence of each (see section 3 for more details of the experimental framework). Table1 lists the results. In the table, the notation quot;sess quot; indicates that the first, third, and fourth bytes are all significant and that the second byte is the sign extension of the third.... In PAGE 2: ... We chose to study the 3-bit scheme, although one could reasonably argue that the 2-bit scheme is better due to simplicity and overhead advan- tages; in any case, the performance results are likely to be very similar for both schemes. Table1 also indicates the high level of compression that... ..."
Cited by 44

Table 1: Frequency of significant byte patterns

in Very low power pipelines using significance compression
by Ramon Canal, Antonio González, James E. Smith 2000
"... In PAGE 2: ... We performed a study with the Mediabench benchmarks [6] to determine the relative frequency of occurrence of each (see section 3 for more details of the experimental framework). Table1 lists the results. In the table, the notation quot;sess quot; indicates that the first, third, and fourth bytes are all significant and that the second byte is the sign extension of the third.... In PAGE 2: ... We chose to study the 3-bit scheme, although one could reasonably argue that the 2-bit scheme is better due to simplicity and overhead advan- tages; in any case, the performance results are likely to be very similar for both schemes. Table1 also indicates the high level of compression that... ..."
Cited by 44
Results 1 - 10 of 103,365