Advanced Architecture Computers

@inproceedings{Dongarra1989AdvancedAC,
  title={Advanced Architecture Computers},
  author={Jack J. Dongarra and Iain S. Duff},
  year={1989}
}
We describe the characteristics of several recent computers that employ vectorization or parallelism to achieve high performance in floating-point calculations. We consider both top-of-the-range supercomputers and computers based on readily available and inexpensive basic units. In each case we discuss the architectural base, novel features, performance, and cost. We intend to update this report regularly, and to this end we welcome comments. 
Programming methodology and performance issues for advanced computer architectures
TLDR
The SCHEDULE package is described which provides an environment for developing and analyzing explicitly parallel programs in FORTRAN which are portable and includes a preprocessor to achieve complete portability of user level code and also a graphics post processor for performance analysis and debugging. Expand
Supercomputer performance evaluation: Status and directions
TLDR
An overview of current practice in supercomputer performance evaluation is presented, methods for characterizing applications and architectures are described, and some approaches for accomplishing their pairing are pointed toward. Expand
The Titan Graphics Supercomputer architecture
The design criteria for the Titan are outlined, and the hardware, architecture, and implementation developed to meet them are examined. The key architectural elements in Titan are described, and theExpand
OR Forum - Perspectives on Parallel Computing
TLDR
It is argued that the lack of uniformity in available architectures is the main obstacle for the breakthrough of parallel computing and formal techniques will have to be developed for the design and implementation of efficient parallel algorithms. Expand
Parallel Computing for Probabilistic Response Analysis of High Temperature Composites
TLDR
This Phase I research identified the multiple levels of parallelism in PCM and the computational strategies to exploit these parallelisms and established the required software and hardware strategies to achieve large scale parallelism. Expand
A Methodology for Performance Evaluation of Parallel Applications on Multiprocessors
TLDR
This methodology consists of an efficient analytic technique which allows us to obtain the average execution time of a parallel application taking into account its structure, defined by a task graph, several task parameters, and the contention caused by the processor-memory interconnection network. Expand
The use of frameworks for scientific computation in a parallel distributed environment
  • R. Armstrong, J. MacFarlane
  • Computer Science
  • Proceedings of 3rd IEEE International Symposium on High Performance Distributed Computing
  • 1994
TLDR
A frame-based approach to scientific computing on distributed computing platforms is introduced to introduce well-defined mappings between the representations of physical phenomena as mathematical structures and the computational algorithms for modeling that phenomena on high-performance parallel computing platforms. Expand
Cost-effectiveness of concurrent supercomputers
TLDR
The concurrency function is used to derive asymptotic estimates for speedup, including Amdahl's Law, and provides a new method for analyzing cost-effectiveness of the processor-memory-communications constituents of a computing system for applications where system cost and execution time are mutually elastic variables. Expand
A portable environment for developing parallel FORTRAN programs
TLDR
An attempt at a short term solution to the transportability problem of producing transportable mathematical software using a package called SCHEDULE, which provides a standard user interface to several shared memory parallel machines. Expand
Tuning the ANSYS kernel LSOLVE for a parallel computer
A solver kernel, LSOLVE, has been added to ANSYS, a finite-element analysis code. LSOLVE is a rank-n update routine that can be implemented efficiently on parallel and vector machines with aExpand
...
1
2
3
4
...

References

Validity of the Single Processor Approach to Achieving Large Scale Computing Capabilities
An instrument for facilitating the calculation of equivalent values includes a plate bearing symbols representing units and dimensions, the plate having a window in which a movable pointer isExpand