Many powerful code optimization techniques rely on accurate information connecting the definitions and uses of values in a program. This information is difficult to produce for programs written with pointer-based languages ...

Programmers often rely on performance analysis tools to provide feedback about the execution of their applications. However, the nature of this feedback is far from satisfactory. Often the feedback is purely descriptive ...

All graph-coloring register allocators rely on heuristics to arrive at a "good" answer to the NP-complete problem of allocation, resulting in suboptimal code. We look at a post-pass to the allocator which removes unnecessary ...

Due to the growing disparity between processor speed and main memory speed, techniques that improve cache utilization and hide memory latency are often needed to help applications achieve peak performance. Compiler-directed ...

To satisfy the quality of service requirements of real-time multimedia applications, networks must employ resource reservation and admission control. In this paper, we describe a new Measurement-Based Admission Control ...

Current simulators for shared-memory multiprocessor architectures involve a large tradeoff between simulation speed and accuracy. Most simulators assume much simpler processors than the current generation of processors ...

Reduction recognition and optimization are crucial techniques in parallelizing compilers. They are used to detect the recurrences in a program and transform the originally sequential code into parallel code. Because of the ...

ATM technology is expected to gain widespread use both in wide-area networks and in high-speed local area networks. The performance of the TCP/IP protocol suite over such networks is of great importance, as it is widely ...

The explosive growth of the Internet, the widespread use of WWW-related applications, and the increased reliance on client-server architectures places interesting new demands on network servers. In particular, the operating ...

The Arnoldi iteration is widely used to compute a few eigenvalues of a large sparse or structured matrix. However, the method may suffer from slow convergence when the desired eigenvalues are not dominant or well separated. ...