by M. F. Sakr, S. P. Levitan, D. M. Chiarulli, B. G. Horne, C. L. Giles
ftp://ftp.nj.nec.com/pub/sakr/ICML97.ps.gz
Add To MetaCart
Abstract:
Machine learning techniques are applicable to computer system optimization. We show that shared memory multiprocessors can successfully utilize machine learning algorithms for memory access pattern prediction. In particular three different on-line machine learning prediction techniques were tested to learn and predict repetitive memory access patterns for three typical parallel processing applications, the 2-D relaxation algorithm, matrix multiply and Fast Fourier Transform on a shared memory multiprocessor. The predictions were then used by a routing control algorithm to reduce control latency in the interconnection network by configuring the interconnection network to provide needed memory access paths before they were requested. Three trainable prediction techniques were used and tested: 1). a Markov predictor, 2). a linear predictor and 3). a time delay neural network (TDNN) predictor. Different predictors performed best on different applications, but the TDNN produced uniformly good results. 1
Citations
|
593
|
Hierarchical Mixtures of Experts and the EM algorithm
– Jordan, Jacobs
- 1993
|
|
150
|
A time-delay neural network architecture for isolated word recognition
– Lang, Waibel, et al.
- 1990
|
|
88
|
Compiler-directed data prefetching in multiprocessors with memory hierarchies
– Gornish, Granston, et al.
- 1990
|
|
71
|
Interconnection Network for Large-scale Parallel Processing
– Siegel
- 1990
|
|
65
|
Introduction to Parallel Computing
– Kumar, Grama, et al.
- 1994
|
|
49
|
Evidence-based Static Branch Prediction using Machine Learning
– CALDER, GRUNWALD, et al.
- 1997
|
|
39
|
The impact of communication locality on large-scale multiprocessor performance
– Johnson
- 1992
|
|
28
|
Multis: A new class of multiprocessor computers
– BELL
- 1985
|
|
10
|
Locality based control algorithms for reconfigurable interconnection networks
– Chiarulli, Levitan, et al.
- 1994
|
|
8
|
Dynamic mapping and load balancing with parallel genetic algorithms
– SEREDYNSKI
- 1994
|
|
5
|
Comparisons of seven neural network models on traffic control problems in multistage interconnection networks
– Funabiki, Takefuji, et al.
- 1993
|
|
5
|
Routing in Optical Multistage Interconnection Networks: a Neural Network Solution
– Giles, Goudreau
- 1995
|
|
4
|
Using recurrent neural networks to learn the structure of interconnection networks
– Goudreau, Giles
- 1995
|
|
1
|
Partitioning and Mapping of Large FEM-Graphs by SelfOrganization
– Dormans, Heiss
- 1995
|
|
1
|
Communication Network Routing Using Neural Nets-Numerical Aspects and Alternative Approaches
– Fritsch, Mandel
- 1991
|
|
1
|
Guest editor of the Special Issue on Neurocomputing
– Habib
- 1995
|
|
1
|
Neural Network Controller for Adaptive Routing in Survivable Communication Networks
– Jensen, Eshera, et al.
- 1990
|
|
1
|
On-Line Prediction of Multiprocessor Memory Access Patterns
– Maggini
- 1996
|
|
1
|
Performance of On-Line Learning Methods in Predicting Multiprocessor Memory Access Patterns
– Sakr, Levitan, et al.
- 1996
|
|
1
|
A Neural Network Cache Controller
– Stigal, Dagli, et al.
- 1991
|
|
1
|
Unsupervised Algorithms for Learning SpatioTemporal Correlations
– Tumuluri, Mohan, et al.
- 1996
|
