A Note on Off-Line Permutation Routing on a Mesh-Connected Processor Array

Kriz̧anc, Danny

doi:10.1142/s0129626491000239

Cited by 5 publications

(3 citation statements)

References 2 publications

(2 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…There are some algorithms pursuing to achieve 2n−2 lower bound of routing steps with the queue size of constant factor at O(1) [9−11] ; while some other algorithms pursue smaller queue size with slightly bigger routing steps, such as 2n + O(n) routing steps and a small constant queue size [12] , 2n + o(n) routing steps and queue size 5 [13] , 2.5n routing steps and queue size 2 [14] , and 2n − 1 routing steps and queue size 4 [12] .…”

Section: Related Workmentioning

confidence: 99%

Multistage Off-Line Permutation Packet Routing on a Mesh: An Approach with Elementary Mathematics

Chiew

2009

J. Comput. Sci. Technol.

View full text Add to dashboard Cite

Various methods have been proposed for off-line permutation packet routing on a mesh. One of the methods is known as multistage routing, in which the first stage is crucial. For the first stage of routing, the previous study normally converts it to a problem of graph theory and proves the existence of solutions. However, there is a lack of simple algorithms to the first stage of routing. This article presents an explicit and simple approach for the first stage of routing based on elementary mathematics.

show abstract

Section: Related Workmentioning

confidence: 99%

Multistage Off-Line Permutation Packet Routing on a Mesh: An Approach with Elementary Mathematics

Chiew

2009

J. Comput. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…For MIMD meshes, Krizanc proved a 2.5n steps off-line permutation routing algorithm with queue size two [151]. Kaklamanis et al presented a 2n&1 steps MIMD algorithm with queue size four [133].…”

Section: Off-line Permutation Routingmentioning

confidence: 99%

Packet Routing in Fixed-Connection Networks: A Survey

Grammatikakis

Hsu

Kraetzl

et al. 1998

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

“…Οι παραπάνω τρεις λειτουργίες είναι σχετικά απλές και έτσι μπορούν να υλοποιηθούν με καλά αποτελέσματα στα περισσότερα διασυνδετικά δίκτυα όπως π.χ το mesh δίκτυο. Λόγω της απλής κανονικής του δομής, το δίκτυο αυτό έχει προσελκύσει το ενδιαφέρον πολλών ερευνητών [152,153,154,155,156,157,158,159,160,161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,176,177,178,179,180,181,182,183,184,185,186,187,188,189,190].…”

Section: επεκτάσεις σε άλλα διασυνδετικά δίκτυαunclassified

Παράλληλες Υλοποήσεις Θεμελιωδών Εργαλείων Της Τεχνολογίας Των Πολυμέσων

Κωνσταντόπουλος¹

View full text Add to dashboard Cite

Εισαγωγή 6 1.1 Τεχνολογία των πολυμέσων -Ανάγκη για αυξημένη υπολογιστική ισχύ 8 1.2 Το βασικό μοντέλο παράλληλης επεξεργασίας 12 1.3 Η οργάνωση της διατριβής 14 2 Παράλληλοι αλγόριθμοι πολλαπλασιασμού πινάκων σε συσχετιστικό μοντέλο επεξεργασίας 2.1 Η αρχιτεκτονική μας 2.2 Το μοντέλο της Συσχετιστικής μνήμης 2.3 Αλγόριθμοι Πολλαπλασιασμού Πινάκων 25 2.3.1 Ο αλγόριθμος στο mesh δίκτυο 26 2.3.2 θέματα προσαρμογής 2.3.3 Οι αλγόριθμοι 2.3.4 Σχετική έρευνα 37Γ then i and / differ at the r-th bit else i and / differ at the (m -l)-th bit όπου χ είναι μία οποιαδήποτε δυαδική τιμή, και φ είναι η πράξη του αποκλειστικού-Ή.for a = 0; i < N; «' + +){ A[i,j] = A[G m (i),G m (j)] B[i,j] = B[G m (i),j] C[i,j] = 0 } (b) Mesh input for 1-GRAY for J = 0; i < N; i + +) { A[i,j} = A[G m (i),G m (j)] B[i,3\ = B[G m {i),G m {3)] C[i,3] = 0 } (a) Mesh input for 2-GRAY forali«,.? €[0,--,iV-l] foralli,j€ [0,---,JV-l] A[G m (i),G m (j)]= A[G m (i),G m ((i + j)modN)] A[i,j]= A[i,G m (j)] A[G m (i),j] = A[G m (i), (i + j) m odN] AUX[i,j] = A[i,fl AUX[i, j] = AUX[i, N-l-j] B[G m {i),G m {j)]= B[G m {(i + j)modN),G m (j)] B[G m {i),j] = B[G m {(i + j)modN),j] for k = 1 to Ν -1 do C[i,j]= C[i,j) + A[i,j]-B[i,j\ A[i,G m (j)]= A[i,G m ((j + l)modN)] *) *) *) *) *) *) *) *) *) *) *) *) *) *) *) *) (* PARTIAL SUMS*) (* WHEN (* ESTIMATING *) *) *) *) *) *) *) *) *) *) *) *) (* EACH PYRAMID*) (* LEVEL *) d: INTEGER; (* THE MAXIMUM *) *) *) *) *) *) *) *) *) *)

show abstract

A Note on Off-Line Permutation Routing on a Mesh-Connected Processor Array

Abstract: We show how to off-line route any permutation of a n X n meshconnected processor array in 2.5n -3 steps with at most 6 packets per processor per time step and in 2.25n +3 steps with at most 8 packets per processor per time step.

Cited by 5 publications

References 2 publications

Multistage Off-Line Permutation Packet Routing on a Mesh: An Approach with Elementary Mathematics

Multistage Off-Line Permutation Packet Routing on a Mesh: An Approach with Elementary Mathematics

Packet Routing in Fixed-Connection Networks: A Survey

Παράλληλες Υλοποήσεις Θεμελιωδών Εργαλείων Της Τεχνολογίας Των Πολυμέσων

Contact Info

Product

Resources

About