Results 1 - 10
of
38,849
Table 11: StreamIt performance results.
2004
"... In PAGE 7: ... We evaluate the performance of RawPC on several StreamIt benchmarks, which represent large and pervasive DSP applications. Table11 summarizes the performance of 16 Raw tiles vs. a P3.... ..."
Cited by 26
Table 11: StreamIt performance results.
2004
"... In PAGE 7: ... We evaluate the performance of RawPC on several StreamIt benchmarks, which represent large and pervasive DSP applications. Table11 summarizes the performance of 16 Raw tiles vs. a P3.... ..."
Cited by 26
Table 11: StreamIt performance results.
2004
"... In PAGE 7: ... We evaluate the performance of RawPC on several StreamIt benchmarks, which represent large and pervasive DSP applications. Table11 summarizes the performance of 16 Raw tiles vs. a P3.... ..."
Cited by 26
Table 11: StreamIt performance results.
2004
"... In PAGE 7: ... We evaluate the performance of RawPC on several StreamIt benchmarks, which represent large and pervasive DSP applications. Table11 summarizes the performance of 16 Raw tiles vs. a P3.... ..."
Cited by 26
Table 11: StreamIt performance results.
2004
"... In PAGE 7: ... We evaluate the performance of RawPC on several StreamIt benchmarks, which represent large and pervasive DSP applications. Table11 summarizes the performance of 16 Raw tiles vs. a P3.... ..."
Cited by 26
Table 11: StreamIt performance results.
2004
"... In PAGE 7: ... We evaluate the performance of RawPC on several StreamIt benchmarks, which represent large and pervasive DSP applications. Table11 summarizes the performance of 16 Raw tiles vs. a P3.... ..."
Cited by 26
Table 11: StreamIt performance results.
"... In PAGE 7: ... We evaluate the performance of RawPC on several StreamIt benchmarks, which represent large and pervasive DSP applications. Table11 summarizes the performance of 16 Raw tiles vs. a P3.... ..."
Table 2: Statistics Dependent on Applications (ratio)
2000
"... In PAGE 2: ...1 GBytes. Table2 summarizes several statistics depen- dent on the application. We identified the packet stream having the same source address, destination address and application (port number) as an individual flow.... ..."
Cited by 9
Table III lists the distributions of the number of streams originated and terminated at the same pair of hosts. It can be seen that the majority of flows (about 85%) contain only one stream and another small portion of them sink large number of streams. For example, in UFL the biggest number of streams contained in the same flow is 3028. This is caused by some new applications that use a range of port numbers or dynamically allocated ports for their sub-transactions. For instance, in BWY, several hosts employ EDONKEY to download files, which chooses 6667 as their destination port number and continually changing numbers as source ports. Previous studies have shown that, long-lived streams carry a high proportion of the total packets and bytes on international links [11][14]. In the calculation of the average packet arriving rate in Section IV, we only consider those
Table 3: Application scenario.
"... In PAGE 9: ... The bulk data transfer is implemented via continuous streaming of data using the same reliable protocol than for the application updates. For the simulation of different loads and priorities of the involved data streams, we suggest the application scenario of Table3 . The timings in Table 3 are indica- tive and multiple state changes are stretched over an interval of several seconds.... In PAGE 9: ...2. Lossless local loop simulation Figure 12 reports the results of the application simu- lation of Table3 , which ran in the local loop setup of Figure 11. From the curve describing the point resolu- tion of the 3D video object, the different application periods, i.... ..."
Results 1 - 10
of
38,849