TCP Vegas: new techniques for congestion detection and avoidance
SIGCOMM '94 Proceedings of the conference on Communications architectures, protocols and applications
Public international benchmarks for parallel computers: PARKBENCH committee: Report-1
Scientific Programming
B(its) T(o the) U(ser): a Communication Benchmark Proposal
B(its) T(o the) U(ser): a Communication Benchmark Proposal
BTU: A Host Communication Benchmark
BTU: A Host Communication Benchmark
lmbench: portable tools for performance analysis
ATEC '96 Proceedings of the 1996 annual conference on USENIX Annual Technical Conference
| Hi-index | 4.10 |
In a distributed environment, CPU speed is only one factorý and not necessarily the most importantýin determining the time an application needs to complete its tasks. In some applications, the dominant factor may instead be communication. Most existing benchmark suites measure just the computational performance of a given workstation or the maximum network throughput under ideal conditions. The authors have developed the BTU (bits to the user) benchmark to take into account both concurrent activities within a workstation (such as CPU and I/O processes that compete for resources) and concurrent activities on the network. BTU can be used to more realistically assess a workstation's communication performance under a particular load. It provides detailed reports to evaluate hardware and software configurations and identify their communication bottlenecks. The BTU benchmark produces results at various levels of abstraction from a single number that characterizes average performance to a report of how all the individual test suite components performed.