Open AccessProceedings Article
Structured memory access architecture
Andrew R. Pleszkun,Edward S. Davidson +1 more
- pp 461-471
TLDR
The authors investigate one method of reducing the Von Neumann bottleneck by reducing addressing overhead and improves system performance by efficiently generating operand requests, making fewer memory references, and maximizing computation and access process overlap.Abstract:
The authors investigate one method of reducing the Von Neumann bottleneck. Program referencing behaviour is determined by analyzing dynamic address traces. The structured memory access (SMA) architecture uses a computation processor (CP) and a decoupled memory access processor (MAP) with special hardware for efficient accessing of program and data structures and for effective branch and loop management. Prototypical SMA machines are compared to conventional Vax-like machines. SMA machines perform significantly better than conventional machines with the same parameter values. The SMA architecture reduces addressing overhead and improves system performance by (1) efficiently generating operand requests, (2) making fewer memory references, and (3) maximizing computation and access process overlap. 8 references.read more
Citations
More filters
Proceedings ArticleDOI
Complexity-effective superscalar processors
TL;DR: A microarchitecture that simplifies wakeup and selection logic is proposed and discussed, which will help minimize performance degradation due to slow bypasses in future wide-issue machines.
Journal ArticleDOI
Essential Issues in Multiprocessor Systems
Gajski,Jib-Kwon Peir +1 more
TL;DR: A suitable classification scheme for comparing these architectures with the aim of determining the predominant importance of new parallel algorithms for enhancement of computer performance is proposed.
Journal ArticleDOI
PIPE: a VLSI decoupled architecture
James R. Goodman,Jian-tu Hsieh,Koujuch Liou,Andrew R. Pleszkun,P. B. Schechter,Honesty C. Young +5 more
TL;DR: The architecture and its implementation ofAbsm~-PIPE (Parallel Instructions and Pipelined Execution) is a research vehicle for high performance VLSI architectures and organizations and makes extensive use of architectural queues.
Journal ArticleDOI
Bottlenecks in multimedia processing with SIMD style extensions and architectural enhancements
TL;DR: The MediaBreeze architecture is proposed, which uses hardware support for efficient address generation, looping, and data reorganization (permute, packing/unpacking, transpose, etc.) and provides a better performance than a 16-way processor with current SIMD extensions.
Journal ArticleDOI
A Simulation Study of Decoupled Architecture Computers
TL;DR: This paper presents a decoupled architecture that splits instruction processing into memory access and execution tasks into two separate sets of instructions, one for accessing memory and one for performing function execution.
References
More filters
Proceedings ArticleDOI
Implementation of precise interrupts in pipelined processors
TL;DR: Five solutions to the precise interrupt problem in pipelined processors are described and evaluated, with results showing that, at best, the first solution results in a performance degradation of about 16%.
Pipe: a high performance VLSI architecture
TL;DR: The pipe architecture (parallel instructions and pipelined execution) is proposed as a research vehicle for studying high performance VLSI architectures and organizations and its planned implementation is described.