DocumentCode :
1672670
Title :
Analysis of cache memory strategies for some image processing applications
Author :
Cernuschi-Frías, B. ; Hamkalo, J.L. ; Pfefferman, J.D. ; Gonzalez, H.J.
Author_Institution :
Fac. de Ingenieria, Buenos Aires Univ., Argentina
Volume :
3
fYear :
2001
fDate :
6/23/1905 12:00:00 AM
Firstpage :
334
Abstract :
Neural networks and image processing algorithms typically use very large amounts of data and usually this data is processed iteratively. Hence, the issue of cache memories for enhancing the processing speed is important. A particularly important model that fits these applications is the simple loop model. Here, the exact solution for the cache memory simple loop model under random replacement is given using an urn model and the theory of Markov chains. The probability distribution is obtained as a quotient of Stirling Numbers of the Second Kind. It is also shown that asymptotically the number of elements in the urns follows a Truncated at Zero Poisson Distribution
Keywords :
Markov processes; Poisson distribution; cache storage; image processing; memory architecture; Markov chains; Stirling Numbers of the Second Kind; Truncated at Zero Poisson Distribution; cache memories; image processing algorithms; probability distribution; random replacement; simple loop model; urn model; Application software; Cache memory; Image analysis; Image processing; Neural networks; Probability distribution; Process design; Signal analysis; Signal design; Signal processing algorithms;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Image Processing, 2001. Proceedings. 2001 International Conference on
Conference_Location :
Thessaloniki
Print_ISBN :
0-7803-6725-1
Type :
conf
DOI :
10.1109/ICIP.2001.958119
Filename :
958119
Link To Document :
بازگشت